I have been doing some testing on Fairings, and pre-coolers

[selfish_meme]

After reading a few comments on the forums I decided to chase them up. Below are the results.

Fairing Bug Test

If you stage the fairings on the ground you get rid of the mass but not the aerodynamic efficiencies: True

Javascript is disabled. View full album

Pre-coolers allow you to go faster and higher while air breathing: Inconclusive

Edit: After fixing the deficiencies in the first test it seems the difference is negligible and can be accounted for by extra intake air. It is better than a fuselage part but if you need more LF it's probably better replaced by the LF tank.

Javascript is disabled. View full album

Edited July 19, 2015 by selfish_meme

[Foxster]

That's interesting stuff. I do, however, remain unconvinced about the first bit.

I did some of my own experimenting...

With fairing

Same craft with fairing deployed before launch

Drag numbers present

What I found is...

For craft with fairing:

dV before launch - 4991

dV at 100km orbit - 2306

dV used - 2685

For craft with fairing deployed before launch:

dV before launch - 5863

dV at 100km orbit - 2379

dV used - 3484

So it took more dV to get the craft without the fairing to orbit. However, if no fairing was used then the mass saving compensates for the drag and overall you end up with more dV by orbit.

You can see from the last picture that drag is being accounted for with the fairing deployed. The craft was also more wobbly, indicating that the air-flow was not even because of the uneven payload.

You also have to remember that fairings (undeployed) have quite high drag themselves and launching a craft with the fairing pre-deployed removes that drag.

Conclusions:

Fairings have high drag and high mass. Quite often you are better off not bothering with one if you have relatively few or low drag parts to occlude. Also the size and shape of the fairing is very important, have a large blunt one and it is heavy and draggy.

I don't know how to work out the break-even point for using a fairing. Best bet is to launch with and without and see which flies better.

I suspect that the 1.0.4 change of halving the mass of fairings did not go far enough. The drag needs halving too.

Edited July 19, 2015 by Foxster

[GoSlash27]

selfish_meme,

To be scientific on the fairing test you'd need a third launch that would test an otherwise identical craft that never had a fairing.

That way you could see clearly if adding a fairing and then deploying it on the pad has any benefit.

Best,

-Slashy

[Red Iron Crown]

That's such an interesting result that I felt compelled to test it, too. I thought that maybe the reduced mass of the rocket with the fairing jettisoned was dominating the increased drag, so I built a slightly bigger rocket for more testing (pic, craft file). My results:

If launched without the fairing installed in the VAB, it performs as one would expect: The fairingless rocket starts off quicker due to its lower mass but is quickly overtaken by the one with the fairing due to drag differences.

If a similar fairing is installed in the VAB but staged away before ignition, the behavior changes: The fairingless rocket still starts off quicker, but now it keeps its advantage and gets further and further ahead of the rocket with the fairing.

Seems pretty clear to me that the aerodynamic benefits of the fairing do not disappear when the fairing is jettisoned. Good experiment, selfish_meme, I've confirmed your results.

[GoSlash27]

Good work, RIC! Useful info there.

I saw another variation of this trick employed in one of my challenges: A rocket was placed inside a Mk2 cargo bay with a cubic o-strut attached to the front.

Cycle the doors before launch, and you have a zero-drag launch vehicle.

Best,

-Slashy

Edited July 19, 2015 by GoSlash27

[Rhomphaia]

Did some testing of my own.

Javascript is disabled. View full album

The main reason Foxster did not see the performance improvement is because after his fairing is jettisoned, those draggy parts now occlude the fairing, so its lowered drag is irrelevant. (also given how fat that fairing is it probably did not reduce the drag anyway). Jettisoning an actual fairing is not helpful, especially if what is in the fairing is the same size as what is below it.

If you want nose cones though, build them out of fairings as long and pointy as you can, then ditch them.

Edited July 19, 2015 by Rhomphaia

[Majorjim!]

So you got further without a fairing? Excuse me while I palm myself in the face.

[Foxster]

A little more testing into the idea of using fairings for nose cones.

Generally what I care about when launching a rocket is: How much dV will I have left when I get to orbit. So here are a bunch of variations of the same rocket to test this.

Numbers are: Launch dV-Remaning dV at 80km orbit=Used dV

Long, tapered fairing

6593-4463=2130

Long, tapered fairing deployed before launch

7855-4533=3322

Nothing at all

7996-4603=3393

Fairing on top of the rocket but fairing never built

7855-4468=3387

Big nosecone

7835-4491=3344

My usual composite nose cone for this size rocket

7894-4563=3331

My thoughts:

Deploying a fairing before launch makes some difference compared to leaving it in place. This is a combination of reduced mass (the major factor) and that drag is reduced too.

The best nose cone on this simple craft to give the most dV by orbit was the composite one with an adapter and shockcone.

The lowest used dV to orbit was the craft with the long tapered nose cone.

The craft with the most dV left by orbit was the one with nothing at all on the top!

Edited July 19, 2015 by Foxster

[Foxster]

Just a bit more testing. This time with a science suite and PB-NUK payload.

Long, tapered fairing

6302-3054=3248

Long, tapered fairing pre-deployed

7349-3995=3354

Nothing at all

7461-4059=3402

So, once again, the craft went to orbit and then had the most useable dV available if no fairing was used at all.

[selfish_meme]

What speeds were you doing, the craft has big draggy wings at the back, what if you go more streamlined (no wings) and faster.

[regex]

Also, are you using MechJeb for consistent launches? Without that the results are suspect.

[Foxster]

What speeds were you doing, the craft has big draggy wings at the back, what if you go more streamlined (no wings) and faster.

Dunno - I'm bored with it now. You wanna try that?

[GoSlash27]

Interesting; the one in the first example required only 2,130 m/sec to achieve orbit? How is that possible? That's less than orbital velocity at 80x80.

Best,

-Slashy

[Foxster]

Also, are you using MechJeb for consistent launches? Without that the results are suspect.

Yes.

(Some more characters).

- - - Updated - - -

Interesting; the one in the first example required only 2,130 m/sec to achieve orbit? How is that possible? That's less than orbital velocity at 80x80.

Best,

-Slashy

Well spotted.

Umm, I think that might be a funny with how MJ is calculating dV. It will be calculating dV including the mass of the fairing at launch. Then that gets thrown away in orbit, significantly increasing the dV available. Something like that anyway.

Probably we should use the pre-deployed fairing launch dV of 7855 to give 7855-4463=3392.

Not sure it matters for the number I'm interested in, which is how much dV the craft has available when in orbit.

[SAI Peregrinus]

Using this craft which has an extremely draggy but light payload, I can confidently say that fairings do NOT shield their contents after deployment.

Picture of craft:

Without the fairing, it flips and fails to reach orbit.

With the fairing, it reaches orbit with ~1200m/s delta-V remaining.

With the fairing deployed on the launchpad, it flips and fails to reach orbit.

And just to be clear, it will reach orbit with no payload.

Your payloads were simply so low-drag that the fairing's mass & drag was greater than that of the payload. If your craft with nothing has the highest delta-V remaining, then you shouldn't be using a fairing for that craft/payload. Stock fairings are for monstrosities with boatloads of struts holding them together, non nice aerodynamic payloads like exposed 2.5m tanks. Procedural Fairings are better, and FAR changes things so that shape matters instead of just adding up the drag of the parts, but this is about stock.

Edited July 19, 2015 by SAI Peregrinus

[Foxster]

Using this craft which has an extremely draggy but light payload, I can confidently say that fairings do NOT shield their contents after deployment.

Picture of craft:

https://i.imgur.com/l4H4Zva.jpg

Without the fairing, it flips and fails to reach orbit.

With the fairing, it reaches orbit with ~1200m/s delta-V remaining.

With the fairing deployed on the launchpad, it flips and fails to reach orbit.

And just to be clear, it will reach orbit with no payload.

Your payloads were simply so low-drag that the fairing's mass & drag was greater than that of the payload. If your craft with nothing has the highest delta-V remaining, then you shouldn't be using a fairing for that craft/payload. Stock fairings are for monstrosities with boatloads of struts holding them together, not nice aerodynamic payloads like exposed 2.5m tanks..

Yup, I'd agree with this.

As I said before, there is some break-even point where using a fairing is useful. I don't think this is reached for most small or even medium payloads nor as a nose-cone with no payload.

If someone can suggest an easy way of judging this with or without help from MJ or KER then I'd be interested.

[GoSlash27]

Foxster,

Tough nut to crack.

The DV on orbit is a misleading figure, since DV <> "efficiency". When designing a stage the question isn't "what stage will leave me with the most DV at the end of the mission" but rather "what is the lightest, cheapest stage that will accomplish the mission".

For example, a Rhino can orbit a .5t payload with tremendous DV left over, but that doesn't mean it's the best choice for the job.

The best choice would be whatever can do the job lightest and/ or cheapest (in this case the 48-7S).

When making this comparison, the important question is " is attaching a fairing and then deploying it before launch more efficient than either leaving it on or leaving it off?".

I would calculate this by payload ratio; mass at orbit/ mass at launch. Whatever method yields the most mass in orbit compared to the mass at launch has expended the least fuel during the launch itself and is therefore the most efficient. If there's fuel left over, then that's mass that could've been additional payload.

I personally wouldn't trust MJ or KER to calculate these numbers, and they're available in the map view anyway.

Best,

-Slashy

Edited July 19, 2015 by GoSlash27

[Dixi]

So far I ignore existence of any fairing or any aerodynamic elements for vertical launches. Works like a breeze :-). Less parts = less mass = more dV saved for orbit.

Most problematic for me since 1.0 are aerobrakes. Like braking in Jool atmosphere.

[Foxster]

When designing a stage the question isn't "what stage will leave me with the most DV at the end of the mission" but rather "what is the lightest, cheapest stage that will accomplish the mission".

Depends what you are doing. If you are doing some challenge where weight or cost matter then yes. Most times though for me I want a craft in orbit that is going to get to where it's going with a good safety margin and that means I want it in orbit with the most dV to spare.

That's all I was showing with those rockets. If someone wants to repeat it to show that using a fairing is the cheapest or lightest way to get a rocket to space then I'd love to see that.

[selfish_meme]

Also, once out of the atmosphere having a fairing counts against you, because it provides no aerodynamic benefit. So the fairing needs to be ejected at the right spot to provide the most benefit. Also the craft with the least weight will start to catch up once in space. So the longer you spend burning in space the the closer the gap will be. I just did another series of tests because I think Fosters had too much wing (which impacts an aerodynamic craft negatively) and I also think Mechjeb will fly the craft differently (limit speed etc) so I just used the prograde SAS. Interestingly the results I got show the unbuilt fairing get an Apoapsis of about 4000 km, the built fairing about the same but with a much more circular orbit with the periapsis of 10,000m. The deployed fairing achieved 9000 km. Can we explain that by weight difference?

Javascript is disabled. View full album

Edited July 20, 2015 by selfish_meme

[GoSlash27]

Depends what you are doing. If you are doing some challenge where weight or cost matter then yes. Most times though for me I want a craft in orbit that is going to get to where it's going with a good safety margin and that means I want it in orbit with the most dV to spare.

That's all I was showing with those rockets. If someone wants to repeat it to show that using a fairing is the cheapest or lightest way to get a rocket to space then I'd love to see that.

Foxster,

Maybe I worded it poorly...

The question you're trying to answer is one of efficiency, and DV is a measure of range (sorta), but not efficiency. To quantify the effect (or lack thereof) of attaching and then deploying a shroud, you have to look at the payload fraction rather than the leftover DV.

Best,

-Slashy

[mikegarrison]

Depends what you are doing. If you are doing some challenge where weight or cost matter then yes. Most times though for me I want a craft in orbit that is going to get to where it's going with a good safety margin and that means I want it in orbit with the most dV to spare.

That's all I was showing with those rockets. If someone wants to repeat it to show that using a fairing is the cheapest or lightest way to get a rocket to space then I'd love to see that.

I can (and just did) launch a ship with seats for six and close to 30000 m/s delta-v in orbit. But it cost 930000. If I don't need 30000 m/s, I can get six kerbals up to orbit for a lot, lot less. Bigger is not always better, unless you need the capacity for something. (I was doing a suborbital flight on the Sun, which takes a lot of delta v.)

[Rhomphaia]

Did some more testing.

I installed mechJeb for consistent flights to orbit, noticed that fairings confuse MJ calculations considerably.

Javascript is disabled. View full album

Note the varied mass displayed by Vessel info and DeltaV info, which changes between VAB and flight.

The miscalculations clear up when the fairing is deployed.

I launched the 2m to a 100km orbit and noted the number of units Liquid fuel used and the DeltaV remaining in orbit. these were the results

No Fairing at all.

Liquid fuel used: 4726 units

DeltaV remaining: 1722m/s

Fairing Base but not built

Liquid fuel used: 4805 units

DeltaV remaining: 1601m/s

Fairing flown to orbit

Liquid fuel used: 4776 units

DeltaV in orbit: 1495m/s (mechJeb vacuum DeltaV display), 1565 (my own calculations). After I jettisoned the fairing in orbit, both MJ and I agreed on 1643m/s

Fairing jettisoned before launch

Liquid Fuel used: 4684 units

DeltaV remaining: 1760m/s

Edit: The Dev Build of Mechjeb seems to have patially fixed this issue. Both displays are wrong in the VAB, but both are right in flight.

Edited July 20, 2015 by Rhomphaia

[selfish_meme]

So there is a definite effect, at least some of the fairings benefits remain and the weight is gone.

[Teutooni]

I confirmed OP's observations about precoolers - they have no real effect.

Test vehicle was this (craft file). Needs tweakscale because I couldn't be arsed to make the wings out of multiple parts.

First run: Whiplash with precooler replaced by empty mk1 fuel tank (mass 0.2t): speed at 17km level flight: 1430 m/s

Second run: Whiplash with a precooler emptied of fuel and air intake closed (mass 0.2t): speed at 17km level: 1428 m/s

Third run: RAPIER with an empty precooler with air closed: speed at 20km level 1684, can just barely make it to 1703m/s by doing a shallow dive from 21 to 20km.

Fourth run: RAPIER with precooler replaced by empty fuel tank: speed at 20km level 1687 m/s, can also barely make it to 1702m/s in a shallow dive.

Fifth run: RAPIER with precooler empty of fuel but air intake open: speed at 20km level 1685 m/s, can make it to 1705m/s in a shallow dive. Absolutely no difference. Intake air from precooler reads 0.0 at those speeds/altitudes.

I did not use any autopilots so the tests might not be entirely valid, but I think it is safe to say the effect is negligible if there at all. Any difference in speed could easily be accounted for by small differences in pitch angle.

Edited July 20, 2015 by Teutooni