Sunday, 30 May 2010

Empirical Testing

I managed to find some time to do some nozzle tests today. I turned a nozzle with a blank end and gave it a 15mm-3mm internal taper.

I then drilled a series of progressively larger holes and measured the range for each hole diameter. I drilled holes in the range 1mm-4.5mm in 0.5mm steps. Here is a graph of the results:

From this I can see that 4mm is the optimum hole size (as I do not have any drills that are not multiples of 0.5mm this is as near as I can get).

However, as the last nozzle diameter was 4.5mm I had to make a new nozzle to get the diameter back down to 4mm.

So here it is, the final design which gives a horizontal range of 40ft with the nozzle at a height of 3ft above the ground.

I've not had chance to test the maximum range with the nozzle angled, but 40ft ain't bad. Next I need to test the maximum range with the nozzle angled and calculate the flow rate.

Sunday, 23 May 2010

Heat of battle

Incredibly as predicted the weather man got it right this weekend and it was a scorcher (for the UK 27C = 80F). As mentioned I managed to create a backpack harness for the tank. As I was rushed, I was unable to take detailed stage photos.

Here is a picture of the tank complete with backpack and newly fitted 3/8" tubing.

I had problems working out a way to create a harness. In the end I bought a cheap backpack and cut the 'bag' part away to leave just the shoulder straps and the part of the pack that rests against your back.

I then fastened this to a 16" wide piece of 10mm plywood. I did this by cutting three 2.5"x16" strips of 15mm pine. I placed the plywood on the bench, centralised the back of the backpack on the plywood then placed the three strips of wood on top, sandwiching the backpack material in between the strips and the plywood. I then drilled 4 holes through all three layers in each strip.

I then recessed each hole in the strips using a 20mm Forstner drill bit. This allows me to ensure the nuts and washers I used to bolt the three layers together did not protrude above the surface of the strips which, when in place, rest on my back.

Here is a close up of the back of the pack. The three strips can be seen in between the two black shoulder straps. Note the recessed holes for the fastening nuts and washers.

Using a router, I put a radius on the strip edges so they would not be unconfortable when wearing the harness.

To mount the tank to the plywood sheet I fastened two 1.25" planks together along their edge, then cut two 4" holes centered at the junction of the two planks. When apart, this leaves semi-circular holes for each tube to fit in.

When clamped together with bolts, this grips the tank quite firmly. Here is a top view looking down onto the top of the tank.

This 'clamp' was then bolted to the ply.

However, when loaded with water the tank is heavy, so I solvent welded two strips of PVC onto the outside of each tube. These rest on the 'clamp' above and prevent the tank from slipping through the holes. This also shows the backpack strip radius.

To fasten the tank at the bottom, I created another piece of wood with semicircular 4" holes and fastened this to the ply. The tank tubes are held against this with a 1" strip of steel as can be seen at the bottom of the left hand tube below.

I need to re-think my nozzle desgin though, as going up to a 3/8" feed pipe seems to have reduced the range. You're still gonna get very wet though.

Here are a couple of videos of me using the gun this weekend with my son and his friends (no sound as it's from CCTV).

Thursday, 20 May 2010

Range Testing

On Wednesday evening and with help from my son I did some range testing. As described in the previous post there were a few tweaks I could do.

First I de-burred the inner nozzle diameter and tested this with the system pressurised to 100psi.

Here is a not very scientific test shot.

I then applied an external taper to the nozzle:

As mentioned in the previous post, the idea is to minimise air turbulence as the water leaves the nozzle.

I fully charged the system with water (I still need to find it's capacity in litres) and pressurised to 100psi.

Here is a video showing range and duration.

I still believe I can improve on this. The current nozzle tapers internally to 3mm then has a run of 15mm to the outside world. I plan to machine a second nozzle that exits at the taper end.

I also plan to upgrade the tube from the pressure vessel from 1/4" to 3/8" pipe...however, when I got home from work tonight I was surrounded by my son and friends who were armed to the teeth with water guns. Never mind the nozzle. This means war.

I've diverted my efforts to make a harness to allow it to be worn 'Ghostbusters' style. Watch this space.

Tuesday, 18 May 2010

Tweaking my nozzle

The tapered reamer I ordered arrived:

This is a low quality hand reamer but was sufficiently sharp to cut a taper into the Delrin. This taper has a 3mm minimum diameter so I shortened the nozzle to minimise the run length of the final diameter, although the nozzle still has about a 1.5cm run of 3.5mm diameter.

Here the reamer has been laid alongside the nozzle with the end of the reamer at level with the end of the taper inside the nozzle.

I tested this today at 90psi and got 50 feet out of it with a 45 degree nozzle angle. This is a definite improvement to a nozzle with a constant internal diameter. However, I am still not happy with the way the stream is breaking up.

I'm going to try the following:

  • Examining the taper inside the nozzle showed that I had left a burr on the nozzle inlet and outlet orifices.These need to be removed.
  • Reduce the run length of the final 3.5mm internal diameter of the nozzle so that the end of the taper meets the end of the nozzle.
  • Taper the outside of the nozzle in case there is some air turbulence..
  • Put an internal chamfer on the outlet orifice.
All these are easy to implement and the weather is forecast to be warm this weekend....maybe I can use it in anger....?

Saturday, 15 May 2010

Nozzle Time

I got fed up of waiting for a tapered reamer I ordered from ebay to arrive and decided to make a start on the tapered nozzle without it.

I started with a 500mm long bar of 22mm diameter Delrin:

I then cut a 120mm length of this, and turned it on the lathe to allow me to cut 3/8" 19TPI BSP thread onto it.

Delrin machines beautifully. I'll be using it more from now on.

I had an unexpected problem when I can to use the 3/8" die I bought. The die holder I have is 1 1/4" but the die is 1 1/2". I looked through my stash of threading tools but I have nothing bigger.

I solved this by holding the die in the lathe chuck and turning the chuck with the chuck key. I was able to hold the Delrin rod by hand to keep it straight as the thread was cut.

I then applied a chamfer to make it easier to start the thread.

Here is the proto-nozzle with 3/8" thread.

The tapered reamer I have ordered has a minimum diameter of 3mm so I drilled a 3mm hole through the nozzle.

Here is the partially finished nozzle attached to the 3/8" ball valve.

And alongside:

Thursday, 13 May 2010


I tried the 1.5mm nozzle at 90psi and it was, as I expected, a complete failure. It didn't manage to project more than 10ft...but:

The new lathe tools, 3/8" BSP die and 20mm Delrin rod arrived today. If the tapered reamer comes I'll be able to make a tapered nozzle this weekend.

Tuesday, 11 May 2010

New Nozzle

I filled a standard nozzle with epoxy resin and drilled it out to have a 1.5mm diameter hole. I was hoping that might increase the range but the more I read about hydrodynamics the more I believe it will actually reduce it.

This is because the narrower diameter will resist the flow of water through the nozzle. I'm going to try it as soon as possible.

Super Soaker Central recommend a tapered design for maximum effect but these are not readily available so I've decided to make one.

I've order some 22mm diameter Delrin bar which is easy to machine and I intend to use a tapered reamer to obtain a conical internal diameter.

Sunday, 9 May 2010


I completed the assembly today and managed to get a few tests in.

The first thing I had to do was fit the Schrader valve. These seem to be very difficult to source in the UK unless they are already attached to a tyre. In the end I found a guy on ebay selling these:

They are intended for Mini Metro shock absorbers but they are ideal for me for three reasons:

  1. They have a standard, wide thread (M12 x 1.75) making it easy to source a tap and as the thread diameter is wide, it should give a secure mount.
  2. It is made of brass and so won't rust.
  3. Cost: Only £3.89

The PVC pipe that I am using has a wall width of about 4mm. I was concerned that mounting the above in a tapped hole might result in the PVC becoming distorted, or the valve becoming insecure. To fix this I decided to mount the valve at the union of the 90 degree elbow joint pipe as the elbow joints have a wall thickness of about 6mm.

Because I am paranoid, I also decided to fit an additional PVC shim from an offcut of pipe to give a total wall thickness of 4 + 6 + 4mm = 14mm.

I made the shim by using a hole cutter to cut a 35mm disc from an offcut of pipe.

However, this pipe is of a daimeter designed to fit inside the elbow but needs to sit on the outside of the elbow. This means that the outer diameter is greater than the diameter of the pipe this disc was cut from. To allow this to sit flush on the elbow I 'baked' the disc in the oven at gas mark 6 for 15 minutes to make the PVC malleable. I then pressed it over the elbow until it cooled and hardened. Here is the shim in place on the elbow.

I then solvent welded the disc to the elbow, drilled it out to 10mm then tapped it. I also filed it flat to allow the Schrader valve to fit flush.

Here is the complete assembly:

It was now time to pressure test the system. I fired up the compressor and pressurised to 30psi. There were no leaks. Now it was time to fit the water outlet port.

I decided to use a pneumatic quick fit connector for the outlet as it does not stick out very far and allows quick (dis)connection of the pipe:

This is one of the connectors I had left over from Halloween and has an 8mm pipe diameter.

I used the same 'shim and cook' method to fit this at the very bottom of the tank. My Grandad left me a varied selection of imperial taps, one of which was 1/4" BSP which is the correct thread for this fitting.

I knocked up a basic trigger using a few pneumatic bits and pieces I got from work when they shut down our electronic assembly department. I got the ball valve from ebay.

Here is a close up of the nozzle and ball valve:

Here is a picture of Josh testing it for me at 80 psi.

I was disappointed by the range. Anyone on the receiving end will get drenched as it dumped 3 litres in about 25 seconds. The stream breaks up much earlier than I was expecting. Examining the bore of the nozzle I noticed it was very rough so I drilled it out to 4mm on my lathe.

This had no effect. I then changed to nozzles of 6mm and 3mm. Neither showed any improvement on range. I upped the pressure to 90 psi and got a maximum range of 32 feet. I think that I need to reduce the nozzle diameter or that there is too much resistance in the pipe from the tank to the nozzle. I've filled a nozzle with epoxy and will drill it out when it's dry.

Range could also be increased by upping the pressure. The PVC tube is rated for a maximum of 131 psi, but I feel a bit nervous about taking it so close to its rated limit.

Friday, 7 May 2010

Riotous Assembly

The screw thread end set I got to serve as the water inlet has a different diameter than the 1" pressure pipe I bought to connect it to the main tank:

To get around this I turned a section of 1" pipe down to fit the end set:

Here it is connected to the watwer inlet 'T' piece:

I glued the entire system together today. I'm going to leave it for 24 hours before pressure testing.

Thursday, 6 May 2010

Sticky Situation

I started using the solvent weld to join the pipe sections together. The solvent weld system comes in two parts: A cleaner and the actual solvent weld. These are the most volatile chemicals I heave ever used. I had to ventiliate my work area three times during this operation.

The speed which the solvent sets at is scary. About 3 seconds. When I ordered I did not realise I was going to get 'very fast cure'. This is an understatement . I only just managed to align the parts in time, and there is no going back once it has set.

Monday, 3 May 2010

Checking the design

Today I tried out the design without glue. You can see the small end set threaded cap in the second picture. This has an internal diameter of 1" so which should allow for pretty rapid refilling. I'm checking my stash of threaded pipe fittings for some appropriate means of attaching a pipe to the bottom of the tank.

Sunday, 2 May 2010

New last!

Halloween has been and gone and although we are nearly half way through 2010 I've decided to start a new project.

My 7 year old son, Josh loves a water fight. So we've decided to make the water gun from hell.
After some research I found this design which I thought would meet my usual criteria of fun and danger.

The design calls for high pressure PVC piping which is incredibly expensive in the UK for some reason. The only people who sell it at a low cost per metre insist upon you buying a crate of 5m lengths.

In the end I got my supplies from koi fish suppliers here. This pipe is rated PN9 which is good for 131PSI in old money.

Here is a picture of the parts I ordered:

The T piece and the 4" to 2" and 2" to 1" reducers next to it are to allow me to connect the screw cap end set (at the top left hand side) via the 1" pipe at the bottom. This screw cap is to be the inlet to top up the water. This differs somewhat from the linked design but as the availability of parts are is limited here, a compromise was the only solution.

I have a feeling I will be requiring some custom fittings for this project, so I dusted off my lathe. It has not been run for 10 years but it's still in running order. It comes complete with thread cutting gears, but only if I want to produce Whitworth threads.

The final part of this project is to be a solenoid valve to control flow. The original design uses a ball valve but this is both cumbersome and slow to respond. I intend to use a small MPU to control a 12V solenoid. If this responds fast enough, I should be able to have several modes selectable with a push button:

  • Single shot.
  • 3 shots rapid.
  • Full auto.
  • Continuous.
This all depends upon the performance of the solenoid I've ordered. They don't publish the operational frequency. I'll just have to give it a go.

Here's a pic of the one I ordered: