Building the Mark IVb
| Pressure Vessel | ||
|
Plumbing The bottles are plumbed together with reinforced plastic hose. Holes are drilled in the centre of the base of the bottle and in the cap. The holes are about the same as the internal diameter of the hose. The hose must be a tight enough fit not to blow out under pressure. To get the hose in, I soaked it in boiling water to soften it enough to be forced through the hole; first the cap, and then the bottle bottom. It is crucial that the edged of the holes are smooth in order to get a good seal. Getting a good seal is a serious pain in the neck, you have been warned. |
 Section through joint in pressure vessel |
|
|
Reinforcement The whole pressure vessel is reinforced with the 'bodies' of other bottles. Each of these are the parallel sided tube left when the top and bottom of the bottles are cut off. Three of these will cover the joints between the bottles and four will cover the actual bottles. These should overlap. In order to get them on it is necessary to crush the bottle a bit. On inflation, the bottle will be wrinkled but as the pressure rises the wrinkle will come out as the sleeve stretches and the load will be spread between the bottle and the sleeve. |
 Section showing reinforcement of pressure vessel. |
|
|
Fins These are cut off from corrugated plastic (e.g. from roadkill estate agents' signs). That is not really that stunning. They are held on with cuttings of bent bicycle spoke as per the pictures. The points to note: - these are inserted at an angle through the 'ribs' of the 'coroplast' so as to give it a grip on the plastic; - there is tape as well to discourage the fins from coming off the spokes; - the ends of the spokes that are against the pressure vessel should be sheathed or padded in some way to prevent the ends from digging into the vessel should the fin strike anything (the ground, for instance!). |
||
 Section through corrugated plastic fins with spokes shown in red. |
 Shape of bent spoke used to attach fin. |
 End view of fin showing spokes (red) lying against side of pressure vessel. |
| Nose and Parachute Section | ||
|
Overview In order to best describe how this system is built it would be as well to have a rough idea how it (usually - no guarantees!) works in practice; more details will appear in the course of what follows. The nose is held on by a line, one end of which runs down to a timer on one side of the pressure vessel and the other end of which is anchored, via a rubber band, to the other side of the pressure vessel. On launch a static line pulls a pin out of the timer. When the timer runs out it will then release the line that runs to the nose and the rubber band on the other side pulls it off. The nose (which unhooks itself from the sprung line) pulls the drogue out of the open end of the parachute section. This then fills and rotates the rocket nose up again and draws the main canopy out. The drogue will collapse and it and the nose will sit on the main chute doing nothing until it lands. I have put photographs of the parachute deployment sequence (taken one grey day during testing at Farnborough) here. |
||
|
The 'structural' part of the nose is the top of a bottle, cut off about so that about half an inch of the parallel sides comes with it. A hole is made at the base of the neck, at opposite sides, to take a sliding hook. The sliding hook is made (in my case) from a cutting of bicycle spoke shaped as shown in the picture: one end has a closed loop and the other has an open hook. From the closed end the line mentioned in the overview runs down to the timer; from the open end runs the line to the rubber band. Things are this way round so that the nose can separate from the rocket. The main function of the sliding hook is to provide a lateral impulse to the nose. Since the lines down from the nose lie flush with the rocket, they cannot, in themselves pull the nose off; the sliding hook is crucial in starting this. |
 Section through nose cone. |
|
|
It works like this: at launch the line are taught and the hook lies fully to the timer side of the rocket. When the timer releases its end, the rubber band pulls the hook across the nose until the closed loop strikes the neck. There is enough momentum in the system then to reliably tip the nose sideways, pivoting on the edge nearest the rubber band. Once the nose is far enough off centre to lift the line off the side of the rocket, the rubber band will pull it round the rest of the way, to the point where that end of the line falls off the open end of the hook. The nose is, by then, clear of the rocket body. "External metal parts!", I hear you cry, but no! The top of a bottle is not a good shape so it has been faired with polystyrene. A tunnel under the polystyrene is cut to permit the hook to slide and where this emerges from the nose, the resulting hole is covered with a flexible fairing of light plastic or heavy tape. The hook should be short enough that neither end can protrude or touch the flexible fairing. It should also not be possible for it to become separated from the nose cone and, hence, remain an internal part. In the event of the rocket hitting anything at speed, the polystyrene will compress onto the bottle top and hook, spreading out as it does so and covering the hook - or at least that has been my experience and in the course of developing this I have flattened many a nose on the ground! The parachute and drogue are carried in the bottom part of a bottle; how much of the bottle depends on how big a canopy you want to carry but it is worth erring on the big side. The parachute shrouds are attached to the bottom of the bottle via a rubber band. The rubber band is not essential, but it should protect the parachute and rocket from some of the jolt if it should deploy at a higher speed than intended. |
||
|
The top of this section is smooth, the bottom of the nose is not. The bottom edge of the nose is the cut off bit of bottle top. In the edge of this are cut a number of tabs, as in the drawings. I use eight but six would probably work; four might be pushing your luck. These are set alternately inside and outside the bottle bottom holding the parachute and so keeps the nose in place. It is important that the nose rests on the tops of all the slots evenly otherwise they may tear when the rocket is fired (remember that a thirty gramme nose cone accelerated at 70 G will have an effective weight of over two kilogrammes!). To further discourage tearing I make a small hole at the top of each slot with a hot nail. The heat causes the PET to melt and shrivel away from the nail, leaving the hole with a thick edge which is much less likely to tear than the unthickened plastic. |
 Tabs on lower edge of nose assembly.  Tabs on lower edge of nose assembly showing cutout. |
|
|
On the side of the nose with the open end of the sliding hook, which will be closest to the rubber band, three of the tabs are partly cut away. The radius of the cutaway is the same as that of the bottle. The middle one of the three will be inside the parachute section when the rocket is fired. When the parachute deploys, the nose will pivot on the tops of the two slots either side of this tab and so the tab will move inwards as the nose pivots. If not cut away as shown the tab may snag on the parachute inside. The parachute section is joined to the top of the rocket (or top of the camera section) using a topped and tailed bottle as a sleeve much as the joints in the pressure vessel are sleeved but in this case some tape might be prudent to prevent the parachute pulling the section off the top. The timer that releases the nose is the clockwork motor from a wind up toy (in this case a walking, glow-in-the-dark skeleton thing). The motor, removed from its donor, consists of a small plastic box with a long metal shaft with a knob at the end (the winder) and one or more short plastic ones the made the toy do whatever it did. Trim these off. Bend the winding shaft at right angles part way along and slide the knob up the shaft a bit (as in the figure). A loop of string/floss/&c. can be hooked over the end of the bent winding shaft in such a way that as the motor runs down the shaft will rotate and release the line. The knob will stop the loop slipping up the shaft. Enough line can be wound round the straight part of the shaft to allow the desired run time before the line is released. The casing of the motor has two holes in it through which one can insert a sliver of bottle into the works so as to prevent the motor running until required. In the case of the parachute timer this pin is connected to a line the other end of which is tied to the frame of the launcher so that the pin is pulled out of timer as the rocket is fired. The easiest way of attaching the timer to the rocket is to glue it to the outside of the pressure vessel or parachute section. For competition purposes a fairing needs to be fitted over the top on account of the metal shaft. Although this is simple and robust, the external timer and the fairing carry a penalty in terms of drag. An alternative method is to site the timer inside the sleeve between the top of the pressure vessel and the bottom of the parachute section. This is rather fiddly both to build and use. I used the former method on the rocket with the camera and the internal system on the cameraless winning rocket. |
||
| Parachute | ||
| The parachutes and drogues are both made out of light(ish) weight black bin bags with shrouds of dental floss. In the NPL competition I used two parachutes: the one on the camera rocket, fired in round one was made from two bags and was possibly a bit big for its intended purpose (photography) and the one used on our winning flight in round two was just silly; but that was the idea. It was made of four bin bags and was (flat) about eight feet across. I you cut the bottom seam off a bin bag you will be left with a short, wide tube. Cutting up one side of this will leave you with a flat oblong sheet. In the drawings, the internal lines mark seams between these sheets. I joined sheets with parcel tape. Drogue: This is essentially a square with the side length being the width of the basic sheet. Small knots (red dots) are tied in the corners of the square and halfway along the straight sides to give it some shape and make it easier to attach the shrouds (blue lines). The vent is simply a round hole in the middle (NB: these are not scale drawings). The lines crossing the vent are to attach this to the inside of the nose cone. |
   Diagrams of drogue and two parachutes used at NPL. |
|
|
Vents: Note that vents are essential in a parachute of any useful size. It must be possible for air to 'leak through' the parachute; if it can only escape around the edges then it will become unstable and is likely to pendulum and/or collapse in use. Several teams at NPL 2002 demonstrated this with otherwise beautifully constructed canopies; these are not even remotely pretty, but they work. Small Parachute: This is very similar to the drogue but is bigger, being made of two 'bin bag sheets' joined along their long sides. The locations of knots and shrouds is the same. The crossed lines in the vent is where the line to the drogue is attached. Large Parachute: The centre of this is very similar to the smaller parachute, the only difference being the in the vents. The extension to the 'chute is made with two additional sheets cut lengthwise and attached along the edges of the small parachute and the corners are cut off as shown. Shrouds (eight here) are attached at the (knotted) corners as shown. |
||
| Camera Section | ||
|
The parachute was the tricky bit; the camera was relatively easy. In this case I used a 'throw away' camera for a number of reasons but by far and away the main one was that used cameras are easy to get for free, which is important if you expect to stuff a few into the ground at high speed. It turns out that they are actually surprisingly robust. They are also reusable, if you are careful; if I had not been confident of this I would not have gone near them as the concept of a throw away camera is disgraceful. They are also small (which gives more scope for aiming it) and light. If you are unhappy about reloading a camera that is not designed to be reloaded, you should be able to find a variety of 35mm plastic instamatic cameras for less than a fiver in a charity shop. They will be slightly bigger and may be a bit heavier but still perfectly adequate. The requirements for flying a camera are quite basic: A camera must be held rigidly enough to survive launch and there has to be some means of triggering the shutter part way through the flight. The parachute will take care of returning the camera to Earth gently. |
 Timer mounted on camera.  Orientation of camera in section. |
|
|
If you have got a camera with a self timer built in, then you can ignore this paragraph; I did not. A clockwork motor, the same as used to release the parachute, is glued to the top of the camera so that its winding shaft, suitably bent, will push the button as it runs down. This is the tricky bit because of the lack of power in the motor; it may take quite a bit of 'fine tuning' to get it right. In my case this system meant that the maximum time it can be set for is about 2-3 seconds which is clearly not long enough for the rocket to reach apogee and deploy the parachute and so the following system was employed. The pin in the camera timer is connected to the line attaching the shrouds of the main parachute to the rocket so that as it deploys it triggers the camera timer, giving a couple of seconds for the main chute to fill and the rocket to stabilise before the camera is triggered. The figure shows roughly how I mounted the camera in the Mk4b. The camera is packed in place in a bottle with polystyrene because that is both light and easy to work with, and it will also offer some protection to the camera if things go wrong. The camera and packing are held together with double sided sticky tape. The camera is angled slightly downwards as it is the ground that is of interest - there is no point in filling half the frame with sky. Holes are cut in the bottle for the lens to look through, for the timer pin, and to allow the camera to be wound on. The requirements in terms of holes and packing will obviously vary depending on your camera. |
||
| Launcher | ||
|
The launcher used here is of Clark cable type (the original cable tie launcher was developed by an Australian water rocketeer by the name of Clark, hence the Clarke cable launcher). The launch tube and basis for the launcher is 21.5 mmf UPVC overflow pipe. Since this is the external diameter of the pipe matches the internal diameter of the neck of a pop bottle, the one is a good fit to the other. In this case I found that the fit was good enough to provide a seal during the pressurization although it does rather depend on the state of the dies used in the manufacture of each; a friend in a different part of the country found that his pipe and bottles did not seal. |
||
|
The lower end of the launch tube is closed off with a bottle cap screwed onto a bottle neck 'glued' to the pipe. I know of no good way of gluing the cut off neck onto the pipe but what I have done is to apply the appropriate solvent for the UPVC pipe to it and then slide on the PET bottle neck. The solvent neither attacks PET nor sticks to it so I am really not sure why this works, but it does (usually). My guess is that the result of sliding the neck onto the softened pipe is that the pipe takes up the exact shape of the neck, providing an airtight seal, and that the imperfections in the inside of the neck provide enough friction to prevent it sliding off. Usually. The valve from a dead bicycle inner tube is fitted through the middle of the cap; it is here that a bicycle pump will be attached. It is important that this valve is good, not only to make pressurization easier but to prevent water escaping from the rocket into the pump. Water and bicycle pumps do not mix well. |
 Section through launcher. |
|
|
Four cable/zip ties are arranged round the launch tube with the fat end uppermost and facing in. This end will hook over the flange on the neck of the bottom bottle of the rocket to hold it on the launcher during pressurization. The cable ties are held in place with a metal hose clamp. This will probably need to be tight enough that it causes the tube to deform. The clamp should be kept far enough away from both the rocket and the end cap that this deformation does not damage the seal in either case. A cylindrical collar is fitted around the cable ties. The diameter is chosen so that, when the rocket is in place it, the collar will hold the fat ends of the cable ties in against the side of the bottle neck so that they cannot slip over the flange; this is how the rocket is retained. The collar is not rigidly attached to anything. To fire the rocket, it is slid down, away from the rocket, allowing the fat ends of the cable ties to ride out over the flange on the rocket. I would recommend that a piece of string be attached to the collar to facilitate remote firing. This not essential but is much drier and you will get a better view of the launch. The launch tube can be clamped to something to hold it up. The something should be wide and heavy for stability. During testing I used the back of my tricycle, and for the NPL competition I used a frame built of scrap steel found in a skip. It is a good idea to use some from of blast deflector as (depending on the state of the ground) the exhaust water can spray a fair amount of mud around. |
||