Rocket body shape

[NFunky]

Hey guys, been a while since I've been here, but man, I love these forums.  Hey all, been a long time since I've been here, but man, I really love these forums.

Anyway, I've been building rockets for a while now, both in stock and with Realism Overhaul.  I've also become fascinated with real world rockets through all this and I've noticed there's one topic I rarely see posted about here that I'm very curious about.

What's the real deal with rocket shape?  Feel free to talk about KSP or real world, even though I know they're quite different.  I'm primarily concerned with multistage stacks and the dimeter of the various stages.  (But I'm also very interested in anything to do with this topic.)

My big question is, what are the considerations when designing rocket's stages as far as diameter?  I've seen rockets like the Falcon 9 or Arianne 5 have upper stages with the same diameter as the lower, while others, like the Saturn V or Atlas V don't.

I would think that, aerodynamically, a rocket that is a consistent diameter with a good nose cone would have less drag than one that tapers, but I'm not positive about this.  Also, in stock KSP, it's much more mass efficient to use smaller diameter upper stages, except that you then need an adapter, fairing, or other inter-stage element.  In real life though, the upper stage shape doesn't have as much effect on mass, so it makes much more sense to go with a cylindrical rocket.

[Foxster]

It depends where you are. When lifting from Duna reducing drag matters little, from Eve it is essential. From Kerbin it is quite important. Let's talk about Kerbin though as I suspect that's mostly what you are interested in. 

You can overbuild your rocket and, as long as the stages aren't stupidly un-aerodynamic, you can usually get your craft into space. If you ignore drag completely though your craft might heatplode on the way up or run out of dV before orbit. 

Generally, if you streamline your parts and their joint pieces you can get a craft up without too much drama. This means a nosecone at the top of each stack, using the same diameter parts downwards or where they get fatter or thinner to use a tapered adapter (preferably one filled with fuel). If they are just tanks then putting a nosecone on the bottom helps too. A few longer stacks is usually more efficient than lots of short ones. 

One biggy to watch for is extra bits stuck on. These can produce a lot of drag. One major culprit is struts. These waste a lot of dV by being heavy and draggy, if placed at the top of your rocket they can even cause the craft to flip. But other bits like experiments, solar panels, RCS jets, etc are draggy and want to be tucked away in a service bay. 

Edited September 24, 2016 by Foxster

[NFunky]

Really good point about the context.  I was mainly talking about Kerbin, but also real Earth (or RSS Earth).  Although, I suppose Eve is actually probably a better analog for real Earth than Kerbin for this topic.  Correct me if I'm wrong, but I'm under the impression that aerodynamic forces are even more extreme IRL than on RSS Earth (with or without FAR).

I've been quite curious why more rockets aren't built as slim cone shapes.  I know the N1 was, but I can't think of any others off the top of my head.  Especially for single-stage-to-orbit vehicles, a cone seems an ideal shape for minimal drag on ascent and maximal or reentry, especially if engines are mounted radially so the bottom of the cone can be one big heatshield.  I've been toying with some SSTOs like this with RSS/RO.

On a slightly different subject, in KSP physics, is there really much difference between a cylindrical rocket (one tank in diameter) vs a composite of multiple cylinders like most Kerbal rockets are.  I mean, is there any advantage to using a single large Kerbodyne tank with four Mainsails mounted under it (on radial attachment points or something) or four orange tanks with a mainsail under each?  Mass wise, I believe these two configurations are identical, but do they differ much aerodynamically?

[Red Iron Crown]

2 hours ago, NFunky said:

On a slightly different subject, in KSP physics, is there really much difference between a cylindrical rocket (one tank in diameter) vs a composite of multiple cylinders like most Kerbal rockets are.  I mean, is there any advantage to using a single large Kerbodyne tank with four Mainsails mounted under it (on radial attachment points or something) or four orange tanks with a mainsail under each?  Mass wise, I believe these two configurations are identical, but do they differ much aerodynamically?

The large Kerbodyne is closer to three orange tanks. Reduced frontal area is the big difference, aerodynamically speaking. A single 3.75m tank has about 11m² of frontal area, while three 2.5m stacks have about 14.7m² together. Frontal area is a huge component in drag calculations.

[Andiron]

2 hours ago, NFunky said:

I've been quite curious why more rockets aren't built as slim cone shapes.  I know the N1 was, but I can't think of any others off the top of my head.

I think that it will put too much mass near the bottom of the rocket and lowering the center of mass, resulting in more instability. The drag force also depends of the area of the shape, the bigger the craft cross-section is, the more drag there is. I can see two examples of cone rocket, the Sputnik (somewhat) and the McDonell Douglas DC-X

The theoretical perfect shape for supersonic craft is called the Sears-Haack body

[ZooNamedGames]

Make your rockets like darts and you'll fly right just about every time  .

[DerekL1963]

3 hours ago, NFunky said:

I've been quite curious why more rockets aren't built as slim cone shapes.  I know the N1 was, but I can't think of any others off the top of my head.  Especially for single-stage-to-orbit vehicles, a cone seems an ideal shape for minimal drag on ascent

Cones are more difficult to design, analyze, and build than simple barrels.  Real life engineering has more tradeoffs and more complex tradeoffs than KSP does.

[Norcalplanner]

56 minutes ago, DerekL1963 said:

Cones are more difficult to design, analyze, and build than simple barrels.  Real life engineering has more tradeoffs and more complex tradeoffs than KSP does.

Agree wholeheartedly. From what I've read, SpaceX uses a single diameter for upper and lower stages both to simplify tooling, and to allow cost-effective transport of the finished rocket. Their rockets are around 3.75m, IIRC, because that is the largest size they can transport by rail.  All their adventures with super-chilled fuel and oxidizer are driven in part by their attempt to wring the most possible performance without increasing the diameter. Their factory is inland and surrounded by development, so they're going to need a different facility, likely coastal, to build the BFR for the MCT.

[NFunky]

1 hour ago, Norcalplanner said:

Agree wholeheartedly. From what I've read, SpaceX uses a single diameter for upper and lower stages both to simplify tooling, and to allow cost-effective transport of the finished rocket. Their rockets are around 3.75m, IIRC, because that is the largest size they can transport by rail.  All their adventures with super-chilled fuel and oxidizer are driven in part by their attempt to wring the most possible performance without increasing the diameter. Their factory is inland and surrounded by development, so they're going to need a different facility, likely coastal, to build the BFR for the MCT.

Very interesting!  That makes complete sense and I'm now wondering if maybe similar considerations go into the Proton or Arianne 5.  I thought upper stage diameter was primarily an aerodynamical choice, but this makes much more sense.

Back to KSP, if I were building an Eve ascent vehicle (which I am), is asparagus staging still the ideal?  Does the sheer fuel efficiency of asparagus staging still trump the drag of radial boosters?  Would it actually be more efficient to build a large stack instead?

[Van Disaster]

5 hours ago, Andiron said:

The theoretical perfect shape for supersonic craft is called the Sears-Haack body

This is true, but the rocket exhaust counts as part of the shape. What I find interesting is how many fairing bases are almost right-angle to the fuselage.

Not sure about 1.2 stock, but if you're RSS-ing and using FAR I think you probably want to have as little cross-section as possible, which means 1 or few large tank > many radially attached small ones. Still, aero drag is a small component of rocket ascent loss whatever you do unless you tow a parachute

[Magzimum]

12 hours ago, Norcalplanner said:

Agree wholeheartedly. From what I've read, SpaceX uses a single diameter for upper and lower stages both to simplify tooling, and to allow cost-effective transport of the finished rocket. Their rockets are around 3.75m, IIRC, because that is the largest size they can transport by rail.  All their adventures with super-chilled fuel and oxidizer are driven in part by their attempt to wring the most possible performance without increasing the diameter. Their factory is inland and surrounded by development, so they're going to need a different facility, likely coastal, to build the BFR for the MCT.

Obviously, SpaceX have one specific route between factory and launch site, and therefore a set list of bottlenecks that their rockets need to fit through... but in a more general picture, the only real bottleneck is often the height of a cargo because it needs to fit under a bridge. This goes for rail or road, and is also true for a lot of other process equipment in chemical industry and other heavy industry.

And I totally agree with everyone here that the tube is so much cheaper than a cone. It pays off to use a little more fuel, and have a slightly less efficient rocket if you can make it much cheaper.

[NFunky]

When designing aerodynamic rockets, I've been very curious as to whether the engine shrouds make any aerodynamic difference.  If I build a nice cylindrical rocket with a Mainsail lower stage and Poodle upper stage, is it really considered a cylinder by the game?  If not, would it ever make sense to use an interstage fairing or are they just too heavy?

[Red Iron Crown]

2 hours ago, NFunky said:

When designing aerodynamic rockets, I've been very curious as to whether the engine shrouds make any aerodynamic difference.  If I build a nice cylindrical rocket with a Mainsail lower stage and Poodle upper stage, is it really considered a cylinder by the game?  If not, would it ever make sense to use an interstage fairing or are they just too heavy?

As far as I know an engine in the automatic shroud is treated as enclosed in a cylinder by the aero model. Interstages are better for enclosing clusters or undersized engines when the automatic shroud isn't the same size as the stack.

[Plusck]

On 26/09/2016 at 5:46 PM, NFunky said:

I've been quite curious why more rockets aren't built as slim cone shapes.  I know the N1 was, but I can't think of any others off the top of my head.  Especially for single-stage-to-orbit vehicles, a cone seems an ideal shape for minimal drag on ascent and maximal or reentry, especially if engines are mounted radially so the bottom of the cone can be one big heatshield.

Cones are actually pretty bad, aerodynamically speaking, unless the rocket engine itself is the exact diameter of the bottom of the cone.

Drag at the back of a shape is just as bad - or sometimes actually worse - than drag at the front. In KSP 1.0.5 (I haven't tested for 1.1.x) it was always slightly better, if you had to choose, to put nosecones or adapter tanks at the back of a droptank (or a step from one diameter part to another) rather than at the front.

[magnemoe]

On 26.9.2016 at 10:41 PM, Norcalplanner said:

Agree wholeheartedly. From what I've read, SpaceX uses a single diameter for upper and lower stages both to simplify tooling, and to allow cost-effective transport of the finished rocket. Their rockets are around 3.75m, IIRC, because that is the largest size they can transport by rail.  All their adventures with super-chilled fuel and oxidizer are driven in part by their attempt to wring the most possible performance without increasing the diameter. Their factory is inland and surrounded by development, so they're going to need a different facility, likely coastal, to build the BFR for the MCT.

You also have the payload who want an large fairing. 
Saturn 5 is also more cone formed than most rockets as it was designed for one payload. 
If you want general payload you want support for large fairings. 

[Van Disaster]

12 hours ago, Plusck said:

Cones are actually pretty bad, aerodynamically speaking, unless the rocket engine itself is the exact diameter of the bottom of the cone.

The exhaust fills the low pressure area at the back of a rocket, so that's not really an issue - straight-sided cones aren't terribly good shapes though.

Given SpaceX's BFR can land itself, why does it matter where it's built? just fly it to CC from the factory.

[Red Iron Crown]

57 minutes ago, Van Disaster said:

Given SpaceX's BFR can land itself, why does it matter where it's built? just fly it to CC from the factory.

Surely you jest. There's an absolute ton of regulatory red tape surrounding a rocket launch, plus final assembly will likely be done at the Cape anyway.

[Van Disaster]

12 minutes ago, Red Iron Crown said:

Surely you jest. There's an absolute ton of regulatory red tape surrounding a rocket launch, plus final assembly will likely be done at the Cape anyway.

Musk was suggesting using it as a cargo ship if he can get that past beaurocrats then he can fly the thing from some remote factory, I'm sure. If he wants several launches a day ( which he's going to have to do if he wants his Mars thing to be practical ) then that red tape is going to have to be reduced somehow.

I don't see either as practical, but one follows the other.

[DDE]

On ‎26‎.‎09‎.‎2016 at 11:41 PM, Norcalplanner said:

Agree wholeheartedly. From what I've read, SpaceX uses a single diameter for upper and lower stages both to simplify tooling, and to allow cost-effective transport of the finished rocket. Their rockets are around 3.75m, IIRC, because that is the largest size they can transport by rail.  All their adventures with super-chilled fuel and oxidizer are driven in part by their attempt to wring the most possible performance without increasing the diameter. Their factory is inland and surrounded by development, so they're going to need a different facility, likely coastal, to build the BFR for the MCT.

The Soviets hit a rail-hard limit with Proton's 4.1 m, too; but Vostochny has a 3.8 m limit. BTW, the Soviets also innovated the superchilled fuel and oxidizer SpaceX toys with.

And yeah, at that size the BFR will have to be built at the launch pad or moved by water. At which point you run into N-1's issues head-first.

[NFunky]

Hey all, sorry to resurrect this thread, but I had another question and I thought it'd be better not to start a whole new thread.

 

All these answers really got me thinking about non-standard rocket shapes.  I've particularly been thinking about lifting bodies.  I'm envisioning a flattened Sears-Haack style shape with engines at the back, mostly filled with fuel/oxidiser.  I'm looking into mods like procedural parts to build this.

 

So I want to use this design to allow wingless, heats-hieldless reentry of a fairly large vehicle.  I'm pretty sure that it'd be able to generate enough lift to let it gently reenter and glide back down.  However, in designing the launch system, I came upon an idea that might let me add even more dV to it.

 

Once in the air, if my rocket has enough lift, do I really need a TWR of greater than 1?  I was thinking, use cheap solids to loft the vehicle, then they drop away and the thing flies more like an airbreathing SSTO would, though purely under rocket power.  You can get a lot of dV if you aren't worried about having an initially positive TWR.  Or at least, this is my working theory.

 

My question is, would this work?  Will I lose so much dV to drag/gravity losses, even with a high drag/lift ratio, that it negates the extra fuel I pack?  I'm thinking of testing this hypothesis with a winged rocket for now, has anyone else tried this?  Also, could I simulate a lifting body with a long Mk2 rocket?

[Norcalplanner]

With enough SRBs, it won't matter if your craft generates any lift at all on the ascent. Plenty of upper stages, both in KSP and IRL, achieve orbit with a TWR well below 1. 

[magnemoe]

12 hours ago, Norcalplanner said:

With enough SRBs, it won't matter if your craft generates any lift at all on the ascent. Plenty of upper stages, both in KSP and IRL, achieve orbit with a TWR well below 1. 

Yes, you need more dV in real world but can manage with less TWR on upper stage as you have an longer time to burn, 
Same is true in KSP if you burn directly for Mun or geostationary orbit.
If you only aim for 70 km attitude you have limited time to burn. 

Its a tricks who is useful if you have an upper stage with low TWR but plenty of dV like an munar or interplanetary ship and you find that you core stage burn out at 50 km and 1200 m/s and an twr of 0.6 or something. 
Nose up, you need to keep Ap ahead of you yes this waste fuel but option is to loose both ship and fuel as you will start going down again. 
keep adjusting angle of attack so it Ap stays ahead, you would also want to raise it to get less drag as you will spend many minutes doing this burn, as speed increase you can lower nose more and more and you end up with having to do an 20 m/s circulation burn. 
Now you need to refuel the craft but that is cheaper than launching an new one.