[Experiment] Advanced Nose Cone Type B is useless?

[Destroyer713]

Background:

As I was designing rockets, a thought about the two different Advanced Nose Cones came to me. Is there really a difference between the two at all? The description of the Advanced Nose Cone Type B implies that the slanted nature of it will cause less drag on radial boosters than an Advanced Nose Cone Type A, but was this really true?

Question:

Does the Advanced Nose Cone Type B offer any advantage over the Advanced Nose Cone Type A?

Process:

Two Thumper SRBs were attached radially to an empty core stage with a Mainsail attached. The SRBs were not thrust limited, and had the Advanced Nose Cones attached at the top of them depending on the test. The rocket was held by launch clamps to avoid any tipping before launch. SAS was activated, throttle was set to max (not that it should have made any difference if it was or not), and the rocket was launched. No manual control input was used, and the SRBs lifted the rocket until burnout. The maximum height and speed was recorded through the F3 screen. This process was repeated 3 times for each nose Cone type. No mods aside from KER were used, and these experiments were performed in stock 1.0.4 Aerodynamics.

Experiment A:

The SRBs were attached to the central tank with a radial decoupler and struts, simulating a typical rocket that is flown with detachable radial boosters.

Javascript is disabled. View full album

Results of Experiment A:

Nose Cone A:

• Average Maximum Altitude: 22,184 m
  
• Average Maximum Speed: 532 m/s
  

Nose Cone B:

• Average Maximum Altitude: 21,560 m
  
• Average Maximum Speed: 524 m/s
  

​

Experiment B:

The SRBs were directly attached to the central fuel tank. No radial decouplers or struts were used. This simulates a typical SSTO rocket where the radial boosters do not detach.

Javascript is disabled. View full album

Results of Experiment B:

Nose Cone A:

• Average Maximum Altitude: 24,579 m
  
• Average Maximum Speed: 559 m/s
  

​

Nose Cone B:

• Average Maximum Altitude: 23,908 m
  
• Average Maximum Speed: 551 m/s
  

​

Overall Conclusion:

The Advanced Nose Cone Type B offers no obvious benefit over the Advanced Nose Cone Type A in stock aerodynamics. All vessels that had the Advanced Nose Cone Type A equipped flew faster and higher than vessels that had the Advanced Nose Cone Type B equipped.

tl;dr: The Advanced Nose Cone Type B creates more drag that its Type A counterpart in all obvious cases in stock aerodynamics. It is a completely useless part in terms of efficiency.

(Squad, can you at least rebalance this part so it's at least on par with its Type A counterpart? It looks pretty damn cool on some designs, so it'd be nice to not have to sacrifice efficiency for aesthetics!)

Edited October 20, 2015 by Destroyer713

[Sgt.Shutesie]

But it looks so good, so we use it anyway.

[Destroyer713]

But it looks so good, so we use it anyway.

This I agree with.

Hopefully this gets enough attention that Squad rebalances this part so it's at least on par with the Type A Advanced Nosecone.

[GoSlash27]

This is all correct and the drag cube file predicts it.

If you really want a dramatic reduction in drag, try using the tail connecter A instead.

Best,

-Slashy

[moogoob]

It's not totally useless. It's better than a bare booster top. Plus looks nicer!

[Alshain]

Not useless, equivalent.

[klesh]

Wait a minute, "B" is the slanted one, yeah? Hrmpf, thats the only one I use.

[pincushionman]

In reality, such a nosecone should create slightly higher drag than its symmetrical counterpart - on its own. I think. But I haven't done aerodynamics in ten years, so take that guess with a grain of salt. It should also create a side load from asymmetric body lift. What it would really be used for would be to control the pressure field around itself more finely, because of how the pressure fields of different components interact with each other. Which is something stock aerodynamics doesn't account for. It may be effective in FAR, though.

[mythbusters844]

It's like you actually expected stock aerodynamics to make sense. Wow.

[dom700]

It's like you actually expected stock aerodynamics to make sense. Wow.

This!

And then, why did you not test it with FAR as well?

[Destroyer713]

It's like you actually expected stock aerodynamics to make sense. Wow.

This!

And then, why did you not test it with FAR as well?

For crying out loud, I was satisfying my own curiosity about if that slanted nose cone did anything to benefit a rocket at all like its description implied. No need to be condescending about it.

The reason I didn't test it in FAR is because I don't use FAR. Simple as that. You're more than free to test it yourself, though. I'm not going to stop you.

Edited October 20, 2015 by Destroyer713

[Agost]

I hope KSP 1.1 x64 will come out soon, so I'll finally start a serious campaign with FAR and a ton of other mods

New aero is fine for most people, but I would have preferred a direct implementation of FAR

[Kerbart]

The results are fairly close and each experiment was done only three times. Not that I'm challenging or disregarding the results, but what was the standard deviation of the results in each test? Just curious how wide the variance was and if there was any overlap, in case the conclusion really should be "not a relevant difference"

I applaud you for doing things the way they should be done: by using cold, hard data.

[peadar1987]

IRL Type A would divert air towards the body of the shuttle, creating a stagnation zone and increasing the drag. I doubt stock aero models this though.

[mikegarrison]

IRL Type A would divert air towards the body of the shuttle, creating a stagnation zone and increasing the drag. I doubt stock aero models this though.

More than that, we know for sure that the stock aero calculation does not model this level of physics.

[TimePeriod]

Bah, its marginal so who gives a damn.

I'll continue to use them unless I have some kind of fanatical need to gain a few m/s.

My designs always have +1k d/v leftover because I don't like my kerbals getting stuck somewhere.

[Destroyer713]

The results are fairly close and each experiment was done only three times. Not that I'm challenging or disregarding the results, but what was the standard deviation of the results in each test? Just curious how wide the variance was and if there was any overlap, in case the conclusion really should be "not a relevant difference"

I applaud you for doing things the way they should be done: by using cold, hard data.

I will admit, this probably wasn't the most "scientific" experiment ever, if only out of laziness on my part. The only reason I did 3 tests for each nose cone and rocket type was because I realized there was some variance, but I didn't want to spend half the day doing 20+ launches for everything.

Unfortunately, trying to plug these values into an online standard deviation calculator (because I can't for the life of me remember the formula) gives me multiple values, and I'm not quite sure which one I'm looking for. I really should have paid more attention when I was learning about this in my college courses...

But taking a quick look at the data, the maximum altitude values never varied by more than 100 m (the differences between the two nose cones was about 500 m), and the maximum speed never varied by more than a single meter per second. Effectively, while the differences in drag between the nosecones are small, the margin of error between the two never overlapped.

[Brainlord Mesomorph]

IRL a flat-sided cone does create more drag than a blunter bullet-shaped cone.

(from my old days with model rockets)

[DunaRocketeer]

I like using the type-B to cap off the similarly tapered Mk3 fuel tanks.

[Red Shirt]

Maybe I am misunderstanding how this all works. I have been using the slanted nose cone to push spent radial stages away from the main ship. Seems to work. I haven't needed sepratrons since doing this and with no exploding collisions. If I am wrong, OK, but it works for me.

[Melonfish]

This is Kerbal Space Program, we don't go for efficiency we literally add "MOAR BOOSTAHS!"

but kudos on the testing.

[cantab]

Now if you want to make a real mockery of the new aerodynamics, put any nosecone on the boosters, then turn it with the rotation gizmo so it faces backwards and is clipped into the booster. See how fast you get.

[ajburges]

IRL a flat-sided cone does create more drag than a blunter bullet-shaped cone.

(from my old days with model rockets)

Only for sub .8 Mach. Go faster and shockwaves produce non-intuitive results. Rockets should go faster.

[Kerbart]

(...)

Unfortunately, trying to plug these values into an online standard deviation calculator (because I can't for the life of me remember the formula) gives me multiple values, and I'm not quite sure which one I'm looking for. I really should have paid more attention when I was learning about this in my college courses...

But taking a quick look at the data, the maximum altitude values never varied by more than 100 m (the differences between the two nose cones was about 500 m), and the maximum speed never varied by more than a single meter per second. Effectively, while the differences in drag between the nosecones are small, the margin of error between the two never overlapped.

A good demonstration of common sense overruling plugging in numbers in formulas -- You're right, obviously the margin of error is much smaller than the differences between the two sets. It never hurts to take a quick look at it, as you did... and to mention that in the test results in case some $&^! like me brings it up

[Psycho_zs]

I use slanted nose cone to hook onto radial decoupler, then attach boosters below. (although you would need to install stock bugfix mod to fix a crossfeed bug. Somebody forgot to disable crossfeed on radial decouplers).