We sort of knew we'd rue the day we asked you lot for your suggestions as to how exactly to launch our Low Orbit Helium Assisted Navigator (LOHAN) spaceplane, and so it turned out to be, as were were buried under a veritable bucketload of ballockets. Click here for a bigger version of the LOHAN graphic What we're looking for …
WTF are we flying here
Don't forget that there is a rocket motor in this and unless you are going to be exotic it will have to be a standard commercial unit. These gunpowder units have a very limited duration, probably not sufficient for an aeroplane to get clear of the balloon. A rocket might stand a chance but will only achieve another 300m or so altitude. The main altitude is provided by the balloon, for this reason the three tier partially inflated solution will be best if that is what you are after.
On the other hand it this is just about proving the technology then you are doing fine.
On the other hand it this is about the fireworks then I would favour a hydrogen balloon with the rocket firing as the balloon bursts passing through the gas cloud. Now that would be AWESOME.
Gunpowder? They have engines fueled by CORDITE?
I assume you mean low-impulse black powder engines. These top out at E power levels; but there are also composite (ammonium perchlorate) engines that go up to (ready for this?) N power. Yep, 500 times more impulse than the largest black-powder motors. They still only burn for about 4–5 seconds, but with an average thrust more like 4 or 5 kilonewtons: one of these will add tens of thousands of feet of altitude, provided we can keep it stable.
My money is on...
...the 'fire it through the burst balloon' option.
If you dangle everything a longish distance below the balloon, and use a tension-based trigger to detect when the balloon bursts, then you have a very simple system. You launch at maximum possible height and the balloon has a bit of time to spread out lots minimising the risk of entanglement.
Add guns or some sort of laser to shoot a hole through the balloon and we're rocking.
Keep it simple, stupid.
Impressed though I am at the ambition of some of the submissions, I can't say they all look a doddle to implement.
With that in mind, I propose a somewhat prosaic waste pipe and curtain track based design.
This assembly would hang, nay, dangle, beneath a single balloon in an inverted T, with the main payload forming the base. A piece of rigid plastic waste pipe of say 32mm diameter would ascend vertically towards the balloon, with the tether at the top end, and with a piece of curtain track curving out and away from the riser, braced for rigidity, terminating at such an angle that a rocket-propelled vehicle leaving said track would clear the swollen balloon.
That angle would obviously depend on how big the balloon gets and how far below the balloon the launch assembly hangs. As already mentioned, the further below the balloon you start, the closer to vertical you can get and still clear it.
Vulture II would sit vertically atop the main payload until launch, in her track and ready to go, and upon reaching altitude she would simply ignite, run the rail and shoot up past the balloon and away.
To my thinking, the known unknown is to what extent the whole assembly would be tilted off vertical as the accelerating V2 rocket (ooh, that's unfortunate, isn't it?) transits the curve. I reckon that could be tested using an appropriately weighted firework with the assembly hanging off a lamp post though. That's assuming the SPB is prepared to spend time launching fireworks along curtain rails attached to waste pipes hanging off a lamp post, of course. I quite understand that as serious journalists you may not wish to get paid for such nonsense.
Looking at Screwfix.com, you'd be out ₤1.85 for a 3m length of 32mm pipe, and ₤7.65 for a curtain track with fittings, and that's before glue, so I grant you it's not cheap. With no moving parts other than the obvious, and no split-second timing required, it just might work though...
I have emailed Lester an appropriately simple schematic of something I like to call the Neatly Implemented Plastic Parabolic Launch Enabler.
Paris, because we don't have one for Lindsay yet.
Well, the doughnut balloon plan kind of has legs, if someone out there can find one,
..how about an inner tube from a large tracter or heavy earthmoving vehicle (such as this: http://www.ausbusiness.net/wp-content/uploads/mining-dump-truck-265x245.jpg).
The upside of one of those, (other than its ability to hold a bit of gas), is that it would already have the necessary valves in place to retain said gas once injected....
This is the way to do it...
Damn straight it isn't sufficient!
Well, not when you can get engines designed for multi-stage use off the shelf anyway. To elaborate on that awful MSPaint-derived sketch, the first stage (the bit with the 'pipe-shaped fin', directional nozzle or whatever you want to stick in the bottom) would just last long enough to eject the rocket+sabots from the tube, flip the whole thing to vertical and get it going fast enough for fins to work. Three seconds perhaps? The second stage would provide the main duration of thrust, and any little D or whatever-class engine you could put in the V2 (harhar) as a third stage would just be a bonus.
Detection of stage separation could be made relatively simple, along the same lines of how a jet-ski detects separation of rider: attach a jack lead between each stage that gets pulled out. For the first stage, the jack lead can double as a signal carrier for nozzle servos. Hopefully the guidance system can sense the jacks being pulled out to go from directed-thrust mode to canard mode, to final burn and coast mode. If the spaceplane's going to autopilot itself back to base then you'll need horizon detection anyway, so use that to detect apogee as well and go from (final burn and) coast mode to glide mode, hopefully at peak altitude.
Of course, I'm not the one writing the software for all this.
Quick (but not cheap)
Why not stick a helium balloon on the top of an Atlas V?
Is the 3 balloons with a launch tube in the middle;
1. It's simple and easy to build.
2. The COG should remain constant or shift slightly downwards during launch.
The main issue with this design is ensuring that Vulture II doesn't get stuck in the launch tube.
No atmosphere = no aerodynamics
I am appalled at the designs with wings that depend on aerodynamic effects for lift. Remember that PARIS released at 89,000 feet, well into the stratosphere. There is insufficient air density at that altitude for a wing to create lift. That's why they use rockets for flight beyond the stratosphere. Rockets have fins to guide it through the lower atmosphere, but the fins have almost no effect once it hits the stratosphere. At that point, the rocket basically only has two forces acting on it, lift (the engine) and gravity. It takes careful design to get a rocket with an engine at the bottom to fly. You have to keep the center of thrust carefully aligned with the center of gravity.
So the LOHAN rocket stage cannot rely on wings during the boost phase. It's going to have to be basically a Congreve Rocket, the engine way up at the top, well ahead of the center of gravity. You could put a rocket on the end of a stick, like a bottle rocket, and it would perform adequately at stratospheric altitudes. The only force you can rely on at this height is gravity, you'll need it to keep the rocket oriented upwards.
Once the boost phase is over, the aircraft can basically tumble through the stratosphere. There is no air to provide lift and no drag that can orient the plane forward. It will need a wing design that has adequate lift, once it hits the denser lower atmosphere, and a good dihedral angle that can automatically correct a roll, and a good tail fin to correct spin. But none of those wing surfaces will have any effect at launch height.
IMHO the best design provided by readers is suspending an angled launch platform at the end of a 50 meter cable. This is enough distance to clear the balloon. Look at the wikipedia entry for Rockoon. It shows a photo of a Deacon Rockoon, it uses this design. It's been tested by real rocket scientists and it works. The wiki entry has a link to JP Aerospace, their Rockoon page has a similar design, and shows an actual launch photo.
No atmosphere, eh?
Of course there's atmosphere up there. A balloon is a lighter-than-air vehicle, thus it requires air to be lighter-than, otherwise no lift.
If there's enough air to hold a balloon up, there's enough air to use for a bit of drag for stabilisation; only a small force is required. Nobody's looking to create aerodynamic lift to hold the rocket up.
Also, the "rocket on the top with no other stabilisation" is a failure waiting to happen: http://en.wikipedia.org/wiki/Pendulum_rocket_fallacy
Flight stabilisation is needed. It's Big Fin or Gimballed Rockets. A rail isn't going to help at all once the rocket leaves the rail.
A boat is a lighter than water vehicle, it needs water to float, but theres no water above it is there?
Holy hardware, Batman!
Nice discussion. There are a couple of failure modes that seem likely enough to warrant fool proof avoidance plans. A vertical launch through the center of a cluster of balloons seems sensible given the craziness of schemes to launch from below the rocket. A system that retards or accounts for the swing of the balloon should be paramount. Look at how much PARIS was swinging. Regarding thrust vectoring, I read the Armadillo Aerospace blog for years as they developed their attitude-maintaining, thrust-vectoring vehicle. An alarming number of sensors fed into a remarkably complex amount of code and it took scores of tries to work it all out. I don't believe for a minute that you'll be able to get software and hardware to work well enough to use vectored thrust, especially as all the sensor inputs, calculations, commands, and reactions have to be done in milliseconds if you're going to get through any significant number of cycles before a black powder model rocket motor burns out. I also think you should use the popping of the balloon(s) as the trigger to launch because if you trigger before the pop by even a little, you'll be throwing away far more altitude than you'll get from any normal sized rocket motor.
Complex perhaps, but I don't think the needs of the rocketry part of LOHAN are all that much. I don't think it needs to go at precisely N degrees from the vertical then fly through a street, stopping at the traffic lights. Simply "somewhere vaguely upwards" would suffice. So maybe launch at 45 degrees and have the rocket guide itself to a point where a few IR sensors are saying "the ground is down there, you're alright"?
Bear in mind that the Vulture 2 will have to be guided by GPS to get back to base, and even with existing autopilots that's not going to be a simple task. That and, well, amateur guided rockets have been done before. It's not too big a stretch to see it being done again for LOHAN, perhaps using some of the same concepts. Plus the onboard footage from a rocket that isn't spinning wildly (as unguided rockets tend to do) would be pretty awesome.
Has none of the dangly designers ever played Thrust? Your rocket will launch at a random angle. There is only one sane solution: put the rocket inside the balloon, attached to the top, and stick a big spike on its nose. Easy.
If there is going to be air at the launch altitude, there is not going to be a lot, and it is not going to provide a great deal of steerage, even at top speed.
Yet, at the same time, you want wings for descent control and range.
Concurrently, if they are not going to provide steerage, they are going to provide drag, thus reducing the final altitude.
Folding wings are clearly too complex and liable to failure.
So why not a flexible wing, like a hang glider, with either rigid or inflatable structural beams. ( I like the inflatable idea).
The wing sits inside the rocket casing until max height, then a small explosion, or the inflation of the tubes, blows off a casing and the wing deploys.
Elementary, my dear Watson.
"Folding wings are clearly too complex and liable to failure."
"The wing sits inside the rocket casing until max height, then a small explosion, or the inflation of the tubes, blows off a casing and the wing deploys."
And your's isn't more complex???
Your's would need inflatable structures to be designed, control valves, compressed air storage tank and this all assumes you can pack said inflatable structures so they deploy correctly.
See parachute packing.
A more realistic alternative:
If you have a wing the full length of the body inline for launch, with a spring loaded pivot bearing in the middle and a catch.
When the burn is completed and the wing is ready to deploy, release the catch and wing deploys, simple, clean, effective.
I think you're both right. In the (awful MSPaint) design of mine on page 7, you can see the idea of a spring-loaded pivoting wing. I can see this working, however it may well be possible that you'd get more lift by having full-length spars (possibly wing-shaped at the leading edge), with some thin nylon or silk fabric forming the rest of a membrane wing.
Advantage is a greater wing surface area, plus the silk/nylon would be strong enough to prevent the spars from snapping open too hard or too far and flying off. The membrane parts of the wings could probably be folded like a concertina within the Vulture 2's body until it's time to deploy. Essentially, once opened it would look like two right-angle triangles with the longest side facing forward.
Also, elaborating on rocket guidance: If it's shown that (say) an 80-degree-angled tube will align the rocket alright without any complex guidance, could Lester at least think of a single-axis anti-roll mechanism? Not guidance as such, but a pretty simple-to-implement way of stopping the rocket from spinning like a top and making any video footage worthless. Just one gyro, the feedback from which is used to directly control fins on the rocket. The hardest part would be calibrating the amount of gyro feedback so that you get a very s-l-o-w roll, without over-corrections that would cause wobbling. Basically, not very hard at all and probably within the range of many hobby rocket enthusiasts.
... using controlled fins for Aero feedback, will also not be that easy to set up. Settings that work fine up to 10k would be totally u/s at 80k. It would need to be tested in a low pressure wind chamber, (assuming that you can find one).
Also with all rail/tube launches, there remains the issue of icing up. It is probably not feasible to have heating elements all of the way along. For simpler systems, depending on the time to altitude, you could maybe use one of those Ski hand warmer packs for a particular component.
I vote for...
...the simple-yet-effective design proposed by Tim Harris.
I vote for...
...a scale model of Scaled Composites' SpaceShipOne, complete with feathering mechanism. That would be cool.
Or, a scale model of the recently-retired Space Shuttle / Orbiter / whatever it's called. That would also be cool.
I was among the first to suggest vectored thrust and completely agree with everyone who said that it will be very hard to achieve, particularly if Vulture 2 has a similar flight-testing program to Vulture 1 (ie none at all). However, it would be cool to try.
Use Fresnel lens (lightweight) focussed onto Peltier junction module in order to split water (carried onboard) into H2 and O2 which is then piped via fishtank tubing into a pair of compressors.
This then fuels the craft in flight, which means that there is no rocket motor and therefore the CAA restrictions should be a non issue.
In the event of a leak it will "fail safe" i.e. the rocket won't ignite.
Once at altitude activate the ion drive using CO2 cylinder filled with argon or xenon as the ion source.
The SR-71 did it
Before getting silly about lack of atmosphere, the SR-71 flew (sustained flight) at about this altitude - 85,069ft is it's record. http://en.wikipedia.org/wiki/Lockheed_SR-71_Blackbird so there's clearly plenty of air around.
SR-71 was going slightly faster
Vulture 2 will be starting from almost stationary relative to atmosphere in around 1% atmospheric pressure.
I rather doubt it'll hit Mach 3 - and even if it does, it will definitely not get anywhere near there before clearing the launch rail!
Screw the Balloons, How about some REAL DIY Rocket Fuel!!!!!!!!!!!
Even standard "black" gunpowder has it's own self contained source of oxygen (either sodium or potassium nitrate). Smokeless powder is Nitrocellulose based. Why do you think a sealed rifle cartridge still fires?
Anyway, you're going about this all wrong. Who needs balloons?
You could take the easy way out and buy Class M solid fuel rocket engines premade and with a proper multi stage design reach 100,000 ft.
But if you are REALLY SERIOUS, how about "rolling your own" rocket motors Mythbuster Style? No, not the Salami and Liquid Oxygen version. Think BIG or GO HOME!
Carbon Black (think copier toner), finely powdered Aluminium (a fine file and a lot of time), Ammonium Perchlorate (better have a liscense) and some latex rubber emulsion mixed in the right (but not disclosed, that you have to figure out for yourselves) proportions is what the Solid Fuel Boosters for most ICBM/NASA designs use.
Add the mixture to a clean cement mixer with a small amount of water. Just remember to make sure that the cement mixer and anything/one that touches it or the fuel, has anti-static protection. (VERY SURE!)
Use an appropriate tube that will contain the pressure, fill with the emulsified fuel and mold a star shaped tapered hole down the center about halfway. Vibrate the bubbles out, orient tube vertically until latex cures. Allow to dry over a significant time in a temperature controlled environment.
Get some high temp porcelain clay and mold an appropriate rocket nozzle that will fit in the end of the tube, fire it in the kiln, allow to cool, examine for cracks and affix with adhesives and mechanical fasteners.
NASA uses big sparklers (literally) to light the Shuttle Boosters because the sparks light the star shaped edges of the fuel evenly. (Very Important Fact, Uneven Combustion is highly undesireable)
Remember, DO NOT UNDER ANY CIRCUMSTANCES RELY ONLY ON THE USE OF RUBBER O-RINGS TO JOIN THE SECTIONS TOGETHER! (From NASA and Thiokol "Lessons Learned" archives)
Voila, cheap rocket engines of any size or thrust you want and no risky liquid hydrogen and oxygen. Obviously, there is a MUCH steeper learning curve making your own and the accompanying risks are (no pun intended) Astronomical.
Make the engines big enough and Vulture II could leave Earth Orbit from a standing start and no balloons would be required.
However, you might want to let NATO, USAF, RAF etc know that a home made, multi stage, ballistic missile with payload, is being launched. For some reason, they get a little touchy about that stuff.
Or if you want to take the easy way out, just appeal to Elon Musk to be the "Topper" on one of his "Birthday Candles".
Oh by the way no state secrets divulged here, only comonly available, somewhat arcane, knowledge. My father (RIP) told me this stuff. He used to work for Thiokol in Utah.
Flame Icon of Course - (Graphic was lifted from warning sticker for Oxidizers, No?)
A few little nitpicks:
"You could take the easy way out and buy Class M solid fuel rocket engines premade and with a proper multi stage design reach 100,000 ft."
Or, hang the lot under a triple triplet of balloons and reach truly insane heights due to not having to push through the troposphere and all that pesky weather. Sorta like what I drew on page 7.
Also I think Lester has ruled out home-brew rocket engines, possibly for the same reason he's ruled out inflating LOHAN's orbs with hydrogen. No matter, as you can get M-class engines pre-made. Stick one in the bottom of a fibreglass tube with fins on, light the thing up and watch it go.
And no, that's the warning for "Highly Flammable". The oxidizer warning is a flaming circle, sorta like this: http://us.123rf.com/400wm/202/261/chas53/chas530904/chas53090400023/4656700-united-states-department-of-transportation-oxidizer-warning-label-isolated-on-white.jpg
(excuse the long URL)
...ice, ice, baby.
A nominal cooling of 1.98 degrees Centigrade per thousand feet of ascent.
Could be a problem with mechanisms.
Jus' sayin'. (Unless someone has already mentioned it, of course).
Ice only forms in the presence of water vapour.
With that in mind, the whole launch vehicle could be stuck in a tube with a latex seal on either end. The air inside can be dehumidified on the ground, and when it's time to fire up the engines, the little rocketplane can burst through the latex.
Also, if LOHAN is the project and Vulture 2 is the spaceplane, what do we call the launch platform and the possible rocket vehicle that Vulture 2 sits in/on? I'm guessing "Falcon" might attract unwanted attention from a certain commercial rocketry outfit, but you get the idea.
Our El Reg designers are going to be hard pushed to develop that in a garden shed.
Get the ship to a speed that its control surfaces work (needs some sums here) BEFORE firing your rockets onboard.
Launch at any horizontal aspect (0-say 45 degrees) and let the wings provide the control system (known technology, off the shelf) when speed is adequate.
Use the payload as a launcher (mortar or a launched stack of rockets on a rod).
If launched on a mortar then Lohan needs to be strong enough that her wings dont rp off.
Wings CAN be folded - forward - on an titanium hinge and cross brace (simple design - not that expensive) and spring loaded to snap into place, (they are going backwards momentum will do it and simple construction.)
What size/weight and height is she to fly from, realistically, so that sums can be done?
1) Weather balloons increase dramatically in size as they rise. As the air pressure drops the volume of the gas inside the balloon increase. Some folks don't seem to have accounted for this.
2) Weather balloons are far from rigid or stiff in any way. They are made from latex if I remember correctly. Having seen one or two launched they squidge around and change shape like a soap bubble in the wind. There's an awful lot of solid, round looking balloons in the designs. I doubt any kind of "shaped" balloon will work as you have to cater for the increase in volume with altitude and deal with the inherent lack of rigidity in the envelope.
3) The couple of weather balloons I have seen launched with weather measuring / camera payloads underneath have had the payloads swing around rather a lot. I suspect there will be a pendulum / oscillation thing going on that has to be damped out to get any platform stable. This could be difficult to damp out at altitude if there is little air to react against.
4) I have no idea what size or make of rocket motor you intend to use but a few things are worth bearing in mind:
4.1) It is illegal in the UK to make "explosives"* thanks to the 1875 Explosives Act**. Moving commercial rocket motors around (Aerotech, Cesaroni) which are bigger than 20Ns total impulse or made from propellant other than black powder (i.e. Ammonium Perchlorate Composite Propellant) requires paperwork from the HSE . Estes black powder motors aren't an issue. However I doubt they would give the kind of oooomph you are looking for.
4.2) Commercially available rocket motors have burn times in the order of a couple of seconds excepting Boost Glider rocket motors but these are difficult to obtain in the UK due to not necessarily being Classified and Authorised by the HSE or CE marked (See Placing on the Market, Storage and Transport of Explosives Regulations or POMSTER / HSE). Essentially if the motor isn't C&A'd and CE marked it cant be bought or moved in the UK. Every EU country has different rules on rocket motors and what goes for the UK probably wont be the same for other countries so watch out for designing in something you cant fly elsewhere in the EU. I suspect your best bet would be a trip to the US but the rules there are different again.
4.3) Rocket motors and explosive charges are much more difficult to light at higher altitudes. Igniting solid motors depends on there being air present to conduct heat between the igniter and the propellant. This isn't as big a problem, to some extent and depending on motor system selected, for motors as it is for any kind of loose black powder based ejection charge which can be very unreliable at altitude.
4.4) There are 3 main types of rocket motor:
Solid Rocket Motors use a solid oxidiser and a solid fuel. Probably, for ease of use, best in this case.
Hybrid Rocket Motors use either a solid fuel and a liquid oxidise or a solid oxidiser and a liquid fuel. As far as I am aware nearly all hybrid rocket motors use the former.
Liquid Rocket Motors use Liquid Oxidiser (O2, N2O, HTP(H2O2)) and Liquid Fuel (H2, Kerosene Paraffin, Alcohol) or a Liquid Monopropellant (HTP(H2O2), Hydrazine (NH4))
While high altitude ignition of a hybrid is feasible I suspect the additional complexity and mass limits choice to a solid motor. Liquid Rocket Motors are fine for Space Shuttles and Saturn 5s but not for LOHAN. We have seen that LOHAN does not necessarily function well when alcohol is present.
5) Boost gliders are a bitch to design, build and get to work in the atmosphere at ground level (search YouTube for Boost Gliders i.e. http://www.youtube.com/watch?v=kbsafWNDiLo). They can and do work but they are challenging.
6) Issac Newton: An awful lot of the designs seem to assume that when your rocket motor lights and the rocket glider launches the platform some how stays put. That there is no drag between craft and launch guide and that for every action there isn't an equal and opposite reaction. As a lot of the platforms are dangling beneath the balloons and are necessarily going to have to be designed to be light even small amounts of drag between launch rail and craft will lead to the payload and launch guide being dragged around by the rocket plane. Icing on the launch guide could have to be accounted for. I have seen even small model rockets take the launch pad with them at lift off. Remember that commercial rocket motors are designed to give a fast hard kick in the trousers to a rocket. Not a glider. Not a long slow burn (a la Saturn 5 take off's and the TV show Salvage 1). On a 1m long launch rail a small rocket on a G Class motor (up to 160Ns total impulse) can be doing several hundred miles per hour before it leaves the launch rail. Obviously this depends some on motor selection, design of the craft, drag.....
7) Active stabilisation is tricky and vectored thrust is tricky too. Both are achievable http://www.rocketeers.co.uk/node/420 for example. But there is a significant mass penalty if you are considering it for a "Ballocket"
8) General rule of thumb at ground level is that a rocket has to be travelling at at least 30mph when it leaves the launch guide for air passing over the stabilising surfaces to be travelling fast enough to work. Below this speed the restoration forces on the fins/wings are insufficient for the flight to be stable. Quite how this translates to higher altitudes and lower air pressures I don't know. For passively stabilised rockets launched from the ground to this kind of altitude they will already have travelled a significant distance through the atmosphere during which time the fins will have worked to keep them going upwards and Newtons 1st law will keep it going in the direction the fins pointed it.
*Rocket motors, or rather solid rocket propellants, don't explode they deflagrate (burn very quickly) but for the purposes of the 1875 Explosives Act solid rocket propellants, regardless of their composition, are explosives.
** You can make small quantities for, I believe the wording is, "no useful purpose." This was included to allow the demonstration of compounds that may have vigorous reactions to students for educational purposes. I understand this to mean: mix up the constituent chemicals in the appropriate (small) quantities, tip out in to a loose unconstrained mound, ignite from a safe distance, watch it go WHOOOF, look suitably impressed. The moment you constrain it or use it to do something you are making it for a "useful purpose". You can make explosives if you have a licensed premises (not an easy licence to get) but then you still probably run in to C&A and CE marking to move anything from the point of manufacture.
Congratulations to Dan Lakey sneaking his "highly detailed sketch" of the male reproductive organs on The Register.
Naturally I vomitted upon seeing this disgusting filth.
One of my projects
One of those concepts is similar to one of the projects I'm working on.
I wish you success in doing this.
The first one is the winner..
that's such an awesome way to launch something into space, it's poetic! So much so, I've made a concept video to prove it.
Epic multi-media sensation, better than Super 8 and Cars 2 put together! (and 10 seconds long)
One of the main criteria should be to keep it simple!
My thoughts after working my way through the ideas and comments are.
Don't bother with folding wings. Many people have mentioned the reduced drag at that altitude, and it gets less as you go higher. Having the wings and tail permenantly deployed will aid the stability.
Why carry a spaceplane and a balance weight? Might as well make two space planes and increase the chances of a successful flight instead of dead weight. This way you could mount them either side of the balloon via a mesh bag around it, with the bonus that simultanious launches will cancel out the rolling motion caused by firing a single rocket off to the side.
It the mesh bag being loose on the ground when the balloon is small is an issue, small velcro ties or something can be used to gather it up, and will pop open as it expands. The avionics package on the base will keep it upright.
The shape and stability of the balloon shape will be much better at altitude when it has settled and has no other forces on it. Also, the mesh bag with a balloon tight inside it will actually form quite a rigid structure (like an inner tube inside a bike tyre).
There are a lot of valid points about the complexity of a guidence system (even one to fly it bak let alone one that will guide it straight up), and the size and power of the rocket required. I fly RC, but that wont help you much here, but you might want to take a tip from the free flight guys and make it glide in circles. If it flys in a straight line from that height, you might have to catch a ferry to go and get it :-)
I feel like designing one myself now :-)
Good luck chaps!
- Review Samsung Galaxy Note 8: Proof the pen is mightier?
- Nuke plants to rely on PDP-11 code UNTIL 2050!
- Spin doctors brazenly fiddle with tiny bits in front of the neighbours
- Game Theory Out with a bang: The Last of Us lets PS3 exit with head held high
- Flash flaw potentially makes every webcam or laptop a PEEPHOLE