Ultimate SSTO guide

[FlazeTheDragon]

So i got nearly 200h in ksp2, most of which was spent ''researching'' ssto technologies.

I feel like i have most mechanics pretty figured out, considering ive made 200t+ lift ssto's, laythe ssto that can get back to kerbin with out refueling (well fairing 36kerbals) and ssto's with over 10k lko d.v in ksp2.

If you want to check out all the tricks i use, and figure out why youre ssto isnt working very well... or at all even, check out my vid:

 

[regex]

So what kind of speeds are you expecting before you hit the 20km ceiling?

[FlazeTheDragon]

Just now, regex said:

So what kind of speeds are you expecting before you hit the 20km ceiling?

Ideally, you want to be traveling at ~1450m/s at atmospheric engine flameout (~20km) tho thats more of a rule of thumb then something you should go for with every single build.

Really light ssto's will be traveling faster and really heavy ones will be traveling slower.

The idea is that you want to gain as much velocity with the highly efficient atmospheric engines before switching over to less efficient vacuum ones.

[Sequence]

I knew the cargo bays were horrible for drag, but didn't know about the semi-workaround you discussed. Good to know. 

Also didn't realise the shockcones were so inefficient when using a small number of rapiers. I'm still stuck in KSP1 habits. Will have to mix things up a bit more.  

Good video, thanks for posting.

[Lyneira]

A nice overview, thank you for making it!

I have a minor nitpick about the trick you mentioned to reduce a rapier's drag, where you state "the things that actually create drag are these little nodes", because it's an bit of misinformation that has unfortunately been very persistent in the KSP community.  There's a nice video explaining the drag cube system that KSP and KSP2 use by @Lt_Duckweed that goes into much more detail about this. Unfortunately we are lacking debug information tools to view the details of part drag in KSP2 like you can in KSP1, but I have done drag tests on the runway that do line up with the theory and used the principles of this to optimize my tiny Whittle-based SSTO.

It's not actually open nodes themselves that create drag. Rather, nodes are tied to one of 6 of the part's faces that get used in the drag cube system, and allow you to occlude some of the (very draggy) rear face of the Rapier by attaching something to it. Some key points:

• Every part you attach to a craft, no matter if it's "behind something else" will experience drag as if it is fully exposed to the airstream, unless:
  • Attachment nodes occlude some or all of the part's drag cube face(s), or:
  • (when they're not bugged) cargo bays and fairings shield it from drag entirely.
• Moving a part with offset tool does not affect the drag that is generated; rotating it will, though, since it will affect the angle of the airstream to the drag cube's faces.
• The area of the part you're attaching should match up as perfectly as possible with the part you're attaching to (same fuselage sizes, basically), to cancel out that face's drag. But, you'll be adding the drag of whatever part you attached instead, so it needs to be pointy enough to make it worthwhile.
• A drag cube face doesn't produce noticable drag if it is facing parallel to the airstream, so there's nothing to be gained from putting nosecones on nodes that aren't facing into or away from the airstream.
• The rapier's rear drag cube face counts as very draggy, because it's mostly a flat plate. That's why attaching a nose cone is worthwhile. Other engines' rear drag cube faces may count as much pointier, so it might take some experimenting to see if occluding their rear attachment nodes helps.

Typed up from my memory of that drag cube video, if I've missed something or got something wrong please do correct me.

 

Edited March 23, 2023 by Lyneira
Fix wording

[FlazeTheDragon]

28 minutes ago, Lyneira said:

A nice overview, thank you for making it!

I have a minor nitpick about the trick you mentioned to reduce a rapier's drag, where you state "the things that actually create drag are these little nodes", because it's an bit of misinformation that has unfortunately been very persistent in the KSP community.  There's a nice video explaining the drag cube system that KSP and KSP2 use by @Lt_Duckweed that goes into much more detail about this. Unfortunately we are lacking debug information tools to view the details of part drag in KSP2 like you can in KSP1, but I have done drag tests on the runway that do line up with the theory and used the principles of this to optimize my tiny Whittle-based SSTO.

It's not actually open nodes themselves that create drag. Rather, nodes are tied to one of 6 of the part's faces that get used in the drag cube system, and allow you to occlude some of the (very draggy) rear face of the Rapier by attaching something to it. Some key points:

• Every part you attach to a craft, no matter if it's "behind something else" will experience drag as if it is fully exposed to the airstream, unless:
  • Attachment nodes occlude some or all of the part's drag cube face(s), or:
  • (when they're not bugged) cargo bays and fairings shield it from drag entirely.
• The area of the part you're attaching should match up as perfectly as possible with the part you're attaching to (same fuselage sizes, basically), to cancel out that face's drag. But, you'll be adding the drag of whatever part you attached instead, so it needs to be pointy enough to make it worthwhile.
• A drag cube face doesn't produce noticable drag if it is facing parallel to the airstream, so there's nothing to be gained from putting nosecones on nodes that aren't facing into or away from the airstream.
• The rapier's rear drag cube face counts as very draggy, because it's mostly a flat plate. That's why attaching a nose cone is worthwhile. Other engines' rear drag cube faces may count as much pointier, so it might take some experimenting to see if occluding their rear attachment nodes helps.

Typed up from my memory of that drag cube video, if I've missed something or got something wrong please do correct me.

 

yah thats a fair point, could of explained how drag actually works a bit better instead of blaming it all on the nodes

• The area of the part you're attaching should match up as perfectly as possible with the part you're attaching to (same fuselage sizes, basically), to cancel out that face's drag. But, you'll be adding the drag of whatever part you attached instead, so it needs to be pointy enough to make it worthwhile.

Even tho thats how it worked in ksp1, ksp2 seems to be different. I havent ran enough dedicated tests to really say, but from the limited testing ive done, it seems like the optimal aerodynamic cone size is more related to the node size now rather then its actual size.

covering the back of the rapier with the more round nose cone used to give slightly less drag in ksp1 but it actually gives slightly more drag in ksp2.

In addition, using the smallest cone possible to cover up the cargobay nodes also seems to be beneficial compared to an appropriately sized cone, tho this might just be a weird bug interaction.

Ive also experimented breafly with covering the back of the poodle with a cone, and once again, the smallest nose cone seemed to be the best, tho i havent tested it enough to really say as i lowkey ditched the concept of adding dedicated vacuum engines to rapier craft.

 

Would be very interesting to see a new ksp2 based series on aerodynamic effects once you figure out how to properly test them (probs mods).

[Lyneira]

Just to clarify any potential misunderstanding, I did not make that drag cube explanation video. I am definitely looking forward to an updated, accurate ksp2 video on how it works now, though!

[twich22]

Thanks for posting this. When the game is optimized enough for me to be able to fly planes, this will come in handy. 

[Gargamel]

Moved to tutorials 

[Lyneira]

I'm back with another post, to hopefully address a misunderstanding about procedural wings. In the video, you mention that thicker wings provide more lift at low speeds. Lift testing with various procedural wing settings was one of the things I've been doing since patch one and, at this moment, that seems to not be the case at all. Thicker wings do increase wing mass, but don't generate additional lift at low speeds.

As a demonstration, here's my lift tester that I've been using on the runway to test wings at 30 m/s.

Using the runway rather than flying makes other factors like craft mass and angle of attack much easier to rule out as the cause of variation between tests. The wings have been attached with one rotation (5 degrees) backwards so they have a natural angle of attack to generate lift.

I then record the lift generated by the wings at 30 m/s on the runway, for a wing thickness of 0.20 and 0.00.
 

Spoiler

For wing thickness 0.20: 10.158 kN

For wing thickness 0.00: 10.367 kN

The angle of attack does vary slightly (about 0.1 degrees) between both attempts because the thicker wings push the nose wheel down a bit more, but compared to the natural angle of attack of 5 degrees this is a pretty small difference. If wing thickness did provide a boost in lift, the first test's lift should have been noticably higher than the second one, not smaller.

Ultimately, the recommendation to minimize the wing thickness is still valid, but I hope this test helps to show that it's worthwhile to minimize thickness (and therefore wing mass) for craft of any speed.

Edited March 24, 2023 by Lyneira
Add recommendation

[Jarin]

9 hours ago, Lyneira said:

Ultimately, the recommendation to minimize the wing thickness is still valid, but I hope this test helps to show that it's worthwhile to minimize thickness (and therefore wing mass) for craft of any speed.

I do hope this will be updated soon. Either make thickness a cosmetic thing only, or give it a purpose. Having it just be a "voluntarily fly worse" slider is just... bad.

[Lyneira]

Just now, Jarin said:

I do hope this will be updated soon. Either make thickness a cosmetic thing only, or give it a purpose. Having it just be a "voluntarily fly worse" slider is just... bad.

Even if isn't planned to affect lift, fuel tanks in wings may get added later (I sincerely hope so) and the thickness should affect the size of the fuel tank!

[FlazeTheDragon]

16 hours ago, Lyneira said:

I'm back with another post, to hopefully address a misunderstanding about procedural wings. In the video, you mention that thicker wings provide more lift at low speeds. Lift testing with various procedural wing settings was one of the things I've been doing since patch one and, at this moment, that seems to not be the case at all. Thicker wings do increase wing mass, but don't generate additional lift at low speeds.

As a demonstration, here's my lift tester that I've been using on the runway to test wings at 30 m/s.

Using the runway rather than flying makes other factors like craft mass and angle of attack much easier to rule out as the cause of variation between tests. The wings have been attached with one rotation (5 degrees) backwards so they have a natural angle of attack to generate lift.

I then record the lift generated by the wings at 30 m/s on the runway, for a wing thickness of 0.20 and 0.00.
 

  Hide contents

For wing thickness 0.20: 10.158 kN

For wing thickness 0.00: 10.367 kN

The angle of attack does vary slightly (about 0.1 degrees) between both attempts because the thicker wings push the nose wheel down a bit more, but compared to the natural angle of attack of 5 degrees this is a pretty small difference. If wing thickness did provide a boost in lift, the first test's lift should have been noticably higher than the second one, not smaller.

Ultimately, the recommendation to minimize the wing thickness is still valid, but I hope this test helps to show that it's worthwhile to minimize thickness (and therefore wing mass) for craft of any speed.

hmmmm, very interesting, ill need to run some tests on this.

I mostly just assumed that a thicker wing would create more lift.... since thats how it should be xD And increasing wing thickness for a couple of my (not very much worked on ) builds that i thickened up the wing a lot for looks did seem to experience abnormally high drag.

Ill run some tests in the future and make a prompt on the vid if i can get the same results.

[Lyneira]

8 hours ago, FlazeTheDragon said:

I mostly just assumed that a thicker wing would create more lift.... since thats how it should be xD And increasing wing thickness for a couple of my (not very much worked on ) builds that i thickened up the wing a lot for looks did seem to experience abnormally high drag.

Unfortunately, testing drag caused by wings is harder than just reading the total drag number. With wings that have a 5 degree angle of incidence, the lift tester craft slows down very fast when cutting the jet engine's throttle, after reaching 30 m/s. This despite the drag numbers from the Aero GUI in the above tests being extremely low (0.266 kN) and the craft being more massive (1.94t, all fuel tanks full, to keep it from lifting off) than the the same test frame without wings, less fuel mass and a test drag payload with slightly higher drag. (1.10t, 0.281 kN) That one doesn't slow down on the runway nearly as fast.