Air Intakes and Spaceplane Design

[Northstar1989]

A while ago there was this excellent discussion on air intakes and drag in KSP:

This is still an important topic (would be even more so if the dev's could give us some larger airbreathing/jet engines, so spaceplanes are actually useful without massive engine-spam!) and I wanted to continue to draw attention to the idea, discuss it, and see if anything has changed.

Also, there were some nuances to Right's graph (re-posted below for convenience) that I don't think really got any proper discussion- and couldn't be discussed there now without nero'ing a very old thread...

Note, for instance, the shape of the Shock Cone Intake performance curve (or lack thereof).

I think many players sub-optimally assumed the most efficient Spaceplane ascents involve  keeping all your engines lit throughout your entire ascent.

However I have increasingly found this is NOT the case-especially with the 2 stage spaceplane designs I have been experimenting with lately (a smaller Spaceplane optimized for high-altitude, high-speed operation rides piggyback atop a larger plane that breaks off.  Awesome in Sandbox/Science, but requires a mod like Flight Manager for Reusable Stages so you can fly the lower stage back to actually be useful in Career...)  Often it is better to have some engines- particularly Ramjet engines- you only ignite at higher altitudes and speeds, keeping your demand for IntakeAir (and Thrust production) relatively flat as you ascend... (this is even MORE true with modded parts like the Air-Augmented rockets from, I think, Mk2 Expansion: which, realistically for a ducted rocket, perform better at high speeds not only in terms of Thrust, but ISP...)

If you have engines you only ignite at high altitude+speed (or simply don't throttle all the way up until you reach high speed/altitude due to heating issues, aerodynamic stability- particularly with dynamically unstable designs that become less stable at higher speeds, or not having your wings rip off due to aero forces in FAR!) then the Shock Intake curve suddenly looks a lot more appealing: note these curves are for constant altitude- the Shock curve ends up being flattened (in terms of rate of IntakeAir production) by reduced air density at higher altitude...

Other things notable:

- The Divertless Radial Supersonic Intakes appear to have the smallest performance-drop of any intake other than the Shock Intakes between Mach 3 and higher speeds (the slope of their curve is much more gradual, even controlling for their lower peak), making them often the second-best choice for high-speed planes (as well as great for fine-tuning *precisely* how much intake you have, so you don't have any more than needed...)

- Engine Pre-coolers have, surprisingly (and unrealistically, given the whole POINT of using them in real life would be high speed+altitude performance) a steeper curve relative to the amplitude of their peak than the Adjustable Ramp Intake (aka the stock RAM-effect intakes).  This makes them more poorly suited for high speed/altitude operations, at least as intakes (again, this is unrealistic- and the dev's ought to rebalance these to make them more useful).  That being said, form-drag (from frontal cross-section mainly) becomes much more punishing at higher speeds, at least in FAR, so they actually do work well at high speed planes- but for all the wrong reasons (in real life, Pre-Coolers aren't even intakes at all, but allow you to cool/compress airflow before it reaches the engines so they "think" they're actually operating at lower speeds/altitudes.  This would be easily simulated in KSP by simply having them decrease the airflow speed and altitude any engines they are connected too "see"- and indeed this is EXACTLY how they used to or still do work in KSP-Interstellar, which included special code to make pre-coolers work realistically: at least in older versions for sure...)  In real life, they would produce a lot of intake Drag (as you slow the airflow more) and provide no direct intake functionality- yet be CRITICAL for a horizontal-takeoff spaceplane ascent...

- On the topic of pre-coolers, again: there has been some mention that they are highly heat-conductive (wicking heat away from engines), yet this is somehow a BAD thing (as it causes them to absorb more heat from the atmosphere).  It seems to me most players don't understand the Stock heat conduction system well, or how to use this properly.  The best parts to attach pre-coolers to (on the other side of the engine) are large, heavy parts with a lot of cross-sectional area (so these parts in turn can pass the heat they absorb from the pre-coolers to other parts).  This is entirely because the Stock heat model assumes an entire part is all at a constant temperature, to make the calculations manageable.  Anyways, this makes good parts to attach Pre-Coolers to things like the long Mk2-Mk1 adapter, the Mk2 Bicoupler, the flat (rear) end of Mk3 parts, or especially large cross-section mod parts with inline 1.25 meter nodes (like the "Stail" to 2.5 meter adapter with shoulders in OPT Aerospace, or the Mk4 Adapters in Mk4 Expansion...)  The parts they are attached to should, ideally, in turn be attached to even larger parts (like a Mk2-3 adapter in front of a Mk2 Bicoupler).  The key is to wick heat away from the pre-coolers as quickly as possible so they can wick more heat away from the engines in turn.  Not that engine overheating is THAT big of a problem in Stock (except for with the NERVA nuclear rockets- a part intake air precoolers would be USELESS for in real life, unless you were air-augmenting them... Or modded nuclear turbojets, like those in Mk2 Expansion- where at least the use of pre-coolers is realistic)

- The Small Circular Intake has a relatively flat curve that LOOKS like it would be well-suited to high-speed operations: but in reality they tend to explode at high speeds, as they have terrible heat-tolerances...

Edited June 13, 2020 by Northstar1989