These are some things that you can do to bottles so that they perform better. The original work for this was carried out by Clifford Heath - the so-called "Super Guppy" (the two-ended bottle modification that is, not Clifford :-).
The nose of the rocket is the most obvious part to modify as this is the part that the air comes into contact with first.
There are various efficient shapes that you can make the nose and these are detailed on the right:
As a matter of interest, the coefficient of drag for Shape 1 is 0.68 while the figure for Shape 2 is higher at 0.80 (17% higher) - most bought water rocket kits have a nose cone of Shape 2 supplied with them.
Powerful though a water rocket is, it is not going to go supersonic therefore we need only to consider Shape 1.
To modify the bottom of a pop bottle, you first of all need to have the connector described on the connectors page. Attach this to the bottle (as described on that page) and pump up a fair pressure - between 1 and 3 BarG.
Unscrew the pump, leaving the hose (with its one-way valve) in place and start to rotate the bottle, holding the nose approximately 9" above a gas ring. After a while, the plastic will have softened sufficiently to form a hemispherical shape, at which point, you should cool it down under a cold tap or in a bowl of cold water.
There is little danger of a catastrophic failure of the bottle during this process as the place where it is most likely to fail is where it is hottest and thinest. When it does fail like this it is with a Pfffff and not a bang. (you can always use the straight sides for fins).
Modifying the shape of the body is similar in many respects to the process of making the nose into a hemispherical shape. This time, however, the idea is a controlled collapse of the body, down to the required conical shape.
Fix the adapter onto the end of the bottle as above and put a little less pressure in it this time. Make sure that the bottle heats up evenly by turning it continuously above the flame. If the bottle is not showing signs of shrinking but is warm enough then loosen the adapter to let some of the air out. You may have to do this a number of times.
Make sure that you keep the nozzle aligned with the axis of the bottle - if you do not, it may always fly in an arc. Again, once you have got it just the way you want it, put it in cold water to make it set.
Practice makes perfect and it may take a few attempts to get it just right.
You must pressure test any rocket that you make (even testing it after you have glued things on as this can weaken the bottle).
Ideally, you should test the bottle with an incompressible fluid (so that in the event of a catastrophic failure, there is only a small amount of enregy in the system and a situation where shards of plastic fly across the room at superluminal velocities is avoided); at a pressure that is higher than the maximum pressure that you are going to use; for a time that is longer than you are going to need.
Fill the bottle with water and screw the adapter onto the end. Pump it up to around one and a third times the maximum pressure and leave it there for five or ten minutes (or five times longer than you are likely to leave it pressurised on the launcher). It should keep its pressure. Look for bulges, cracks and so on.
Stick some fins on, weight the nose and you are read to go.
Why it works
There are two factors to consider with a modified bottle: the flow of air around the outside of the bottle; and, the flow of the propellant on the inside of the bottle.
in a, the air starts to move back to fill in the void only a short time before it needs to be there. Therefore, the flow rate towards the axis of the bottle is high as the flow is concentrated at the rear end of the bottle.
In b, the air starts to move back towards the centre almost as soon as it has finished moving away from it. The flow lines in diagram b show that the decompression is far gentler in b and this means that less energy is lost. This fact also shows up as a reduced coefficient of drag.
have fins removed for clarity
In diagram c, we can see that, like the air flow around the outside of the bottle, the acceleration of the water towards the centre (and therefore the nozzle) happens only at the end.
Moving water this quickly looses energy and contributes to a higher back pressure than with d, which has a smoother velocity profile. Here, the water is moved towards the centre gradually and therefore the pressure drop is much less.
have fins removed for clarity
Two of the following diagrams are from ready made bottles.
The Normal bottle is a standard 2 litre bottle without modofications. The Velocity profile shows that virtually all of the acceleration occurs within the last 6 cm of the bottle and apart from the concentration of flow at the neck, most of the rest of the acceleration occurs within about 1 cm.
The Soda bottle is an almost conically shaped bottle.The Velocity Profile for this shaped bottle shows that the acceleration between 14 and 4 cm is a fairly smooth curve.
The Home-made bottle, is one that I made earlier (so to speak). Getting a good shape is an artform if you do not have moulds or other specialised equipment. The Velocity Profile shows that this is between the conical and unchanged bottles.
The above diagrams show how the internal and external flows are much more energy efficient in the case of the bottle in b and d (the Home-made bottle and the Soda bottle) compared with the unmodified bottle in a and c (the Normal bottle).
The best shape with regard to making the velocity profile uniform is a mathematical concave shape but this has the unfortunate quality of being quite low in volume (as is the Soda bottle - it is only 1 litre) - ideally we want the rocket to hold enough propellant and air for a respectible flight. Even the conical shape is fairly low on volume - the real bottle seems to be a good compromise between volume and efficiency.
Rockets made from modified bottles or the conically shaped bottles travel higher and faster.