Question on intakes and their parameters

[LordFjord]

Ahoy,

a small question to the modders and non-modders out there.

What are the airbreather-relevant parameters of an air intake?

From the wiki (shock cone intake):

MODULE
{
  name = ModuleResourceIntake
  resourceName = IntakeAir
  checkForOxygen = true
  area = 0.012
  intakeSpeed = 12
  intakeTransformName = Intake  
}

RESOURCE
{
 name = IntakeAir
 amount = 0.4
 maxAmount = 0.8
}

To me, the relevant parameters to support an airbreather engine are the area and intakeSpeed parameters from the MODULE itself.

The RESOURCE part is only how much air it can store, which is pretty much irrelevant (unless for "canned air" operations).

I am searching for a good parameter to measure the effectiveness of an intake.

area * intakeSpeed = intakeVolume / sec

Would that make sense?

It would lead to the following list:

Shock Core Intake = 0.144

Ram Intake = 0.1

Radial long intake = 0.025

...

I also could not find much info on how the intake effectivity reduces with increased altitude. There are no parameters for this in the part descriptions (or I could not find any).

Another question: if my theory is correct, then a single ram intake would generate as much air intake as 4 radial long intakes. Is this also valid when flying on the edge of space?

Or is it an urban (kerbal) legend that air intakes have individual effectiveness regarding altitude (or air density)?

To me it looks as if it does not matter what types of intakes you use as long as the area*intakespeed result is the same.

This is for both some general knowledge about how KSP works and also to find a better algorithm for the mod I am working on.

Thanks ahead,

Fjord

edit: will test this myself tonight, but maybe someone already knows the answer.

[LordFjord]

Well, I did a quick test with a simple SSTO, 2 turbojets, 2 ram intakes as one configuration and 2 turbos with 8 radial long intakes as another configuration. I applied the intake-jets normalisation of my mod after building the jets, so each jet got assigned the same number of intakes.

I tried to follow a similar ascent path.

The two flights were nearly identical. The 8 long intakes flamed out a little bit later but at teh same time ), but that might as well been from the "nearly" identical flight paths. So far, the 2 Rams behaved very similar to 8 long radial intakes.

Gonna try to make a table and test all kinds of intakes. I suspect that all behave the same way, regardless of altitude.

Edited November 24, 2014 by LordFjord

[bdito]

It would be so much easier if they just gave a CFM value for each intake.

[LordFjord]

Here is a test I made, a rather simple plane, 1st placed the Ram Intake and the left turbojet, then placed the 4 long radial intakes and the 2nd turbojet. From the numbers, the airflow for boh engines should be the same, and indeed the flameouts happened very very closely. It dragged the plane a bit to the left, then SAS managed to balance it on its own.

Here is a table with some data:

Name				area		intakeAir		intakeVolume	intakeVolume*10
Shock Core Intake		0,012		12			0,144		1,44
Ram Air Intake			0,01		10			0,1		1
Circular intake			0,008		10			0,08		0,8
XM-G50 Radial Air Intake	0,006		10			0,06		0,6
Structual Intake		0,0025		10			0,025		0,25
Mk1 Fuselage Intake
Engine Nacelle			0,005		10			0,05		0,5
Radial Engine Body		0,005		10			0,05		0,5

I added the last column as an easier measurement/comparison unit - it worked out nicely with the well known Ram intake as reference.

THe structural engine nacelle (without fuel) has no values for airintake, it only works as an intakeair "storage".

So yes, I am now pretty sure that when it comes to intake volume, all intakes behave the same regardless of altitude. I remember reading some maybe outdated advice that the XM-G50 Radial air intakes stop working at high altitudes.

Well, I have not tested those yet in a direct comparison, but I can say that the rams and long radial intakes have the same behavior regarding altitude.

I will now try to find a way to let this knowledge flow into my mod

[GoSlash27]

LordFjord,

This is the million dollar question, and I don't think anyone has gotten it pinned down yet.

Here is a test that shows there are differences, though exactly *why* remains a mystery.

http://forum.kerbalspaceprogram.com/threads/97615-0-25-KSP-air-intakes-compared

I haven't found any correlation between the specs in the config files and the results of this test.

And yes, the idea that the XM-G50 "stops working at high altitudes" is indeed an old wives' tale.

Finally, keep in mind that the most efficient SSTO designs will rely on using as little o2 and rocket as possible, which means achieving or even exceeding orbital velocity on jet thrust alone. It's not enough to just hit 32KM up and then make up the rest with rockets. That will get you there, but your resultant SSTO will wind up larger and heavier.

Thanks for starting this thread. I'll be keeping an eye on it!

-Slashy

[LordFjord]

I haven't seen those intake comparison experiments yet - damned its hard to stumble upon stuff in this forum as there is SOOO MUCH good info scattered around

Thanks a lot for the link, I will take a detailed look at it.

Looking at those testruns in the 1st post, I think the radial intakes get away that good because two of them have a very good intakevolume value, better than a single RAM.

I'll try to include mass and drag values into my table.

Those experiments rerun with identical intakevolumes would be interesting. 1 RAM vs4 long radial intakes or 5 circular intakes vs 4 rams and so on.

I'm sure there is a "best" intake intakevolume/mass/drag wise.

[GoSlash27]

Looking at those testruns in the 1st post, I think the radial intakes get away that good because two of them have a very good intakevolume value, better than a single RAM.

I would've thought so as well, but then the structural intakes perform relatively poorly in the same tests, and have the same intake volume. Even equalizing the intake area between them and the XM-G50 isn't enough to put them on equal footing.

There's something odd going on that the math doesn't readily suggest...

Best,

-Slashy

[LordFjord]

Look at the table in the post above, I took those values from the wiki (might be wrong there but I relied on them so far):

I would've thought so as well, but then the structural intakes perform relatively poorly in the same tests, and have the same intake volume.

The XM-G50 Radial Air Intake beat the structural intakes by over 2 times when it comes to intake volume. You would need ~5 structural intakes to have an approximately similar amount of intakevolume as 2 of the radial air intakes.

2*0.06 = 0.12

5*0.025 = 0.125

5 radial intakes should have the same performance as 12 structural ones.

I wish Squad would simplify the resource flow for intakeair - and simply divide the total amount of intakevolume equally to all airbreather engines. That would instantly make all airhogging and some magical build order rules obsolete.

[GoSlash27]

Something I'm trying to puzzle out now that's worth sharing:

https://github.com/numerobis/KSP-scripts/blob/master/jets.py#L122

The XM-G50 Radial Air Intake beat the structural intakes by over 2 times when it comes to intake volume.

You mean intake area? My info shows the intake volume to be equal. My info also shows the intake area of the XM-G50 as .04 rather than .06.

But my point is that they actually conducted that test, and the structural intakes still lagged behind the others despite the massive cumulative intake volume. So that clearly can't be the overriding factor. The best performers had high volume and the worst performer also had high volume.

Puzzling...

Rank/Type/ area-mass ratio/ area / mass / capacity

1 XM-G50(2) .320 .008 .025 1/2

2 Shock Cone .440 .012 .027 .4/.8

3 ram air 1.00 .010 .01 .2/.2

4 Circular .727 .008 .011 .2/.2

5 Structural(2) .238 .005 .022 1/2

Best,

-Slashy

Edited November 26, 2014 by GoSlash27

[LordFjord]

Hmm, I will need to check this when @home, something seems to be off, maybe the wiki.

Anyway, this should be easy to find with logging a few more lines, then we'll get the data directly from the game.

[GoSlash27]

Hmm, I will need to check this when @home, something seems to be off, maybe the wiki.

Anyway, this should be easy to find with logging a few more lines, then we'll get the data directly from the game.

I'd like to have a definitive answer on that also.

I got my numbers from the config files for the parts, but I can't be certain I got the *correct* parts because I have duplication of parts due to the patch and I'm not 100% certain which is the latest 'n' greatest.

Best,

-Slashy

[LordFjord]

Here is the raw data from my log dump. I could not find any property that displays the intakeair value from the part's config, only a maxintakespeed, which is always 100 in the editor. I dumped the area/mass value as well. It shows the ram as the most effective intake from that aspect.

There were no properties matching the intakeair value, only airflow and airSpeedGui, which were both as expected = 0.

[LOG 19:38:00.830] SF: Intake: shockConeIntake, mass: 0.025, area: 0.012, area/mass: 0.48
[LOG 19:38:00.831] SF: Intake: ramAirIntake, mass: 0.01, area: 0.01, area/mass: 1
[LOG 19:38:00.832] SF: Intake: CircularIntake, mass: 0.01, area: 0.008, area/mass: 0.8000001
[LOG 19:38:00.833] SF: Intake: airScoop, mass: 0.01, area: 0.006, area/mass: 0.6
[LOG 19:38:00.834] SF: Intake: IntakeRadialLong, mass: 0.008, area: 0.0025, area/mass: 0.3125
[LOG 19:38:00.835] SF: Intake: radialEngineBody, mass: 0.15, area: 0.005, area/mass: 0.03333333
[LOG 19:38:00.836] SF: Intake: nacelleBody, mass: 0.15, area: 0.005, area/mass: 0.03333333

So the area values are correct in my table, however the intakeair is giving me a headache. So much for my intakeVolume/sec theory...

Damnit.

Might fall back to only looking at the intake area values.

[numerobis]

My info also shows the intake area of the XM-G50 as .04 rather than .06.

In 0.25, they are 0.06 area. In 0.23 they were 0.04 area. I don't know when it changed exactly.

The size of the IntakeAir tank attached to the intake largely doesn't matter.

Each physics timestep, the intake generates air. If there's more than fits in the tank, the excess is wasted. You can reduce the loss by using a smaller timestep (changing the "max physics deltaT" setting). Or you can ignore the problem because it only happens at low altitude when you have more air than you know what to do with anyway.

Fun trick: from the runway, select a radial intake. It will have about 0.6 IntakeAir in its tank, maybe less if you have a fast computer. Now hit the '.' key to accelerate time. The intake will fill up. Hit ',' and see the intake half-empty itself again.

[pellinor]

There was also a bug in the drag equation for air intakes. Something about a number which is multiplied twice, which results in drag amounts that don't make sense. Does that sound familiar to anyone? Is this still an open bug?

I can't find the reference at the moment but it probably should be mentioned here because it has a big influence on the usefulness of different intakes. The resource gathering part should be unaffected.

Edited November 26, 2014 by pellinor

[bitslizer]

Here is a test I made, a rather simple plane, 1st placed the Ram Intake and the left turbojet, then placed the 4 long radial intakes and the 2nd turbojet. From the numbers, the airflow for boh engines should be the same, and indeed the flameouts happened very very closely. It dragged the plane a bit to the left, then SAS managed to balance it on its own.

http://i.imgur.com/H0CjFUY.jpg

Lord... were you assuming the the left engine would consume the air from the RAM intake while the right engine would consume the structual intake?

IIRC intake air is a global resource, meaning the engines would consume them wherever they are found... so the right engine can be consuming from the ram intake also.

Last engine to be mounted onto the craft get the least priority on resource therefore the last engine added to a craft will always flame out first.

http://forum.kerbalspaceprogram.com/threads/64362-Fuel-Flow-Rules-%280-24-2%29

http://forum.kerbalspaceprogram.com/threads/97615-0-25-KSP-air-intakes-compared

Edited December 12, 2014 by bitslizer

[NathanKell]

That's correct.