How do air intakes work in KSP?

[farmerben]

Is there a guide a air intakes for KSP?

It seems that the biggest difference between the different intake parts is "base effective speed" , what does that mean?

When I right click on air breathing engines I almost always see 100% of propulsion requirements met, or the engines cut off completely due to lack of air.  I never see 90% of propulsion requirements met, so even through trial and error I'm unable to figure out the minimum amount of air intake I need.  I do notice thrust of my engines varies with speed and altitude, but I don't know how to apply this information to decide whether more intakes, fewer intakes, or different intakes would improve performance.  

[bewing]

Once you are going more than 30 m/s, there is no real difference in base effective speed. Any intake will work. The only time that makes a difference is taxiing. Try putting a small circular intake to feed a wheesley, and you will see.

The main difference between the intakes is at very high speed. The intakes choke off at different max speeds. Somebody did some graphs of intake performance with speed, a year or two ago -- your best bet is to search for that.

Or to just try it yourself.

 

[Laie]

On 7/4/2018 at 6:26 PM, farmerben said:

I never see 90% of propulsion requirements met

Try a Rapier and a Shock Cone Intake, clamped down for a static firing test -- it gets to 91% or thereabouts.

In first approximation, intakes pick up air by intake area times air speed. If that formula was followed strictly, zero airspeed would lead to zero intake air... the "base speed" is a means to work around that issue.

Most intakes start to choke at high airspeeds: even though you get faster and faster, there's less air coming in. This is most pronounced with the "subsonic" intakes, but even supersonic ones show that behaviour (somewhere beyond mach 4, so I guess it hardly matters in everyday operations). The sole exception is the shock cone intake, which won't choke no matter how fast you're going.

[4x4cheesecake]

I guess the graph bewing is taling about is this one:

It is created for KSP 1.05 but seems to fit anyway or at least you get an idea about the performance of different intakes. Also, it shows pretty much the behaviour described by Laie.