homemade nozzles

[tim jones]

I'm currently in the procurement phase for my first homemade water gun. In searching for nozzles, I see McMaster-Carr (the tinkerer's friend!) has a few 1-1/2" dia. fire hose nozzles, my local hardware store has a few fixed nozzles with 3/4" garden hose thread, and not much else. The commercial nozzles appear to be optimized for purposes other than water warfare. Threaded caps with drilled orifice holes are a convenient alternative, but they generate lots of internal friction that degrades performance. I am therefore considering the costs and efforts involved in making my own conical nozzles. If anyone has tried this already, could someone direct me to the thread?

The approach I'm considering now is to thermal-form PVC pipe with a two-part mold, a technique inspired by the method spudders use to make flared-base rockets. The internal mold (mandrel) is just a wood rod that fits the pipe ID closely, and has been smoothly tapered to a point. The external mold is a conical tube of sheetmetal, like the bottom part of a metal funnel. To make the nozzle, I would insert the tapered wood rod in a section of PVC pipe, with the pointed end of the rod protruding a bit past the end of the pipe. I'll need to come up with some way to clamp the rod in place, but that should be easy enough. The end of the pipe would be submerged in boiling water for 20-30 seconds to soften the plastic. The hot end of the pipe would then be forced into the sheetmetal cone, thus forming the nozzle shape. The pipe and molds would be plunged into cold water to set the plastic. Finally, the molds would be removed and the nozzle outlet would be drilled for a precise ID.

I see at least three possible complications with this method:
(1) The hot plastic may adhere to the mandrel and resist removal, or leave a rough internal surface. A light coating of high temperature grease should prevent sticking. The wood surface pores can be filled with epoxy for a smoother finish, or a smooth mandrel could be turned out of aluminum.
(2) The hot plastic may buckle when inserted into the female mold. If it tends to buckle axially, then the mandrel can be shifted forward for closer support of the plastic wall, and the nozzle can be formed in several increments. If it tends to buckle circumferentially, then the time in boiling water would be reduced, and forming pressure increased. If it tends to buckle radially, then the length of pipe heated would be reduced.
(3) The conical wall may thicken inward radially, rather than extruding forward. If so, the inerior of the cone wall may be thinned with a drill between incremental molding iterations.

Can anyone think of other potential complications (and, hopefully, solutions) to this method? Can anyone think of an easier way to form nozzles?

[Ben]

I considered making some conical nozzles with some sort of CNC or stereolithography equipment available to me at my university, but I haven't done that yet. That's not exactly homemade though.

On a whim I just searched McMaster-Carr for cones of some sort and can find none.

The 1 1/2 inch fire hose nozzle performed very well. McMaster-Carr also sells some smaller brass conical nozzles for garden hose that I do not recommend. The inside of the nozzle is extremely rough and will decrease performance unless you sand-blast it smooth. Combine that with the cost and weight and I'd rather avoid the brass conical nozzles.

For those who don't know the flare method tim mentioned, check out this: http://www.spudtech.com/content.asp?id=26

I never have done that method, so I don't know much about the potential complications. What you've said makes sense.

The easiest way to make a nozzle might be to pour plastic into a mold manually. I just did some Google searches for stuff like "diy plastic injection molding" and came up with these links:

http://www.cnczone.com/forums/showthread.php?t=9961
http://www.lindsaybks.com/dgjp/djgbk/inject/index.html
http://www.micsaund.com/2006/03/13/p...-diy-projects/
http://www.nhsouth.com/crafts/workbench/injector.htm
http://www.cnczone.com/forums/showthread.php?t=17150

I didn't examine the links thoroughly, but this is a start for DIY plastic injection molding I suppose.

[Silence]

The mold would work provided you have a mold to start from.

I would suggest turning a rod on a lathe to get the shape you're looking for. Obviously that requires access to a lathe capable of molding whatever material you use - metals in particular would require high temperature and strength tools.

The firehose nozzles are a good choice if you have a water gun large enough to use them. They come as small as 1", but even that may be too large. Perhaps you could find a metal funnel somewhere that you could epoxy onto the barrel.

[tim jones]

@SilentGuy: I was planning on turning the mandrel on a "getto lathe" - a drill press modified for this purpose. I've turned wooden items this way, but not metal. Others have told me that small steel items have been turned in this manner, with patience.

Can you quote a source for a 1" firehose nozzle?

@Ben: I picked up a 6" aluminum conical nozzle with garden hose thread yesterday at the local hardware store. The interior is very rough (looks like it was step-drilled), so is probably similar to the small ones sold by McMaster-Carr. The one I picked up is anodized purple, my wife's favorite color. I'll see if I can turn out some of the roughness.

As for injection molding, that sounds like a project I should stay away from until I get a few other projects out of the garage. I have some experience with fiberglass-epoxy molds, so I may try that approach instead.

Side note: I found some 1/2" diameter metal tubes with big rubber handles on a discarded foosball table. These should serve nicely as pump shafts.

[Silence]

Whoops, I was reading the wrong dimension. McMaster-Carr actually carries plastic and brass fire hose nozzles with 5/16", 1/2", and 7/8" apertures. The 5/16" might be a good size for most water guns.

Go to their catalog page 230. In the first chart, you'll see the conical nozzles, labeled D and E (brass and plastic, respectively). They carry each one in two different thread specifications; I think NST is compatible with most PVC fittings.

Anyway, with a lathe, it sounds like your custom nozzle would turn out fine. And since you don't need a particularly strong metal for channeling water, you probably won't need to test the limits of the drill press/lathe.

I believe Ben's conical nozzle did come with a burr on the inside or somewhere that could have used some sanding. But it was one of the plastic ones; I'm guessing the brass nozzle would be rather smooth.

What material do you plan to use for the pump shaft? 1/2" Schedule 40 PVC has an ID larger than .5", but you may need to lathe pump rod to get the O-ring to fit. A 3/4" pump would make it harder to reach high pressures. Other materials may offer sizes in between.

[Ben]

The 5/16 inch nozzle that SilentGuy just mentioned is part number 6484T22 on McMaster-Carr, which costs $22. The price is a little steep. It's probably rough on the inside too, but you never know until you buy it.

I believe Ben's conical nozzle did come with a burr on the inside or somewhere that could have used some sanding. But it was one of the plastic ones; I'm guessing the brass nozzle would be rather smooth.

There might have been one or two burrs that I cleared off and I might have posted about that, but it hasn't stuck in my mind so it wasn't a major problem. My red plastic nozzles are smooth all around. The brass one was rough on the inside.

@Ben: I picked up a 6" aluminum conical nozzle with garden hose thread yesterday at the local hardware store. The interior is very rough (looks like it was step-drilled), so is probably similar to the small ones sold by McMaster-Carr. The one I picked up is anodized purple, my wife's favorite color. I'll see if I can turn out some of the roughness.

Interesting. My brass one is just plain rough on the inside. It doesn't appear step drilled at all. It's like brass sandpaper, but it's coarser than any sandpaper I've seen.

As for injection molding, that sounds like a project I should stay away from until I get a few other projects out of the garage. I have some experience with fiberglass-epoxy molds, so I may try that approach instead.

Molding anything is definitely more involved than other methods, but I just wanted to throw the idea out. I have no experience in molding anything myself, but someone here might.

[Silence]

Wow, I guess I was completely wrong about the brass. I'm just wondering though - is the surface similar to that of brass ball valves without the threads? I have seen some really coarse surfaces on those.

The plastic would definitely be easier to remove the burrs from and polish. It's just a shame McMaster-Carr's supplier didn't bother to use a cleaner method of producing nozzles.

[Ben]

I haven't seen the inside of soldered brass ball valves so I can't make a comparison. But if it's really coarse, it should be similar.

[Silence]

I'm not sure the insides are coarse, but I do know the outsides are. I'm guessing it just comes from the way they machine the valves.

[tim jones]

@Ben: Thanks for pointing out the 1" brass nozzle for $22, I missed that one. When I originally serched McMaster-Carr for nozzles, I just glanced at the $83 price tag on the 2-1/2" nozzle and didn't even look at the rest of the brass models, figuring the cost would be out of reach. I'll still take a crack at the homemade nozzle since I'm cheep, but its nice to know there are options.

I picked up some 1" SDR 21 (200psi) PVC pipe for this project over the weekend, since Schedule 40 pipe looked a bit thick in the walls for the first try. I also turned part of an old shovel handle down to 1-1/8" for the mandrel blank. It now has a nice sliding fit in the SDR 21 pipe. Next, I need to find or make the tapered conical female mold, and cut a matching taper on the mandrel. No luck so far in finding one. Does anyone have an idea for a "found" female mold, before I start looking for my tin snips?

On a related note: Does anyone have small quantities of latex rubber tubing available for purchase? Eternal 46845 and I are both looking for 3/8" ID x 3/4" OD (McMaster PN 5234K53), 3/4" ID x 1" OD (McMaster PN 5234K58), and 1" ID x 1-1/2" OD (McMaster PN 5234K62). We are having dificulty finding these materials locally, and would prefer not having to order 5 feet of each material from McMaster for $77.40 (plus shipping and tax, of course).

[6061]

The brass is probably sand cast and that's what produces the rough finish. If you have a lot of time on your hands you can always get a round file and try alot of elbow greese. I don't why you couldn't make a simple mold out of wax and just use epoxy resin to make whatever shape you want. Once it cures you can just melt the wax out with hot water.

[Ben]

Some cardboard or foam cones from craft stores might be good to start with. If they're not strong enough you could use them as a mold for something stronger and then fill that in. Wax would be easier to remove though, so give that a try.

I have some of the LRT you're looking for at my parents house, but I'm at college so there's not much I can do. I have the 3/8" ID and the 1" ID tubing. You'll probably have to buy the minimum.

If minimums become a problem I could buy some tubing in bulk and resell it. Let me know if you'd be interested in that.

[Silence]

How much of a taper do you want? How long can the nozzle be? What size is the aperture?

6061, does sand casting use the high melting point of sand? I think I see what you mean, and the surface of these brass piece does have the right texture. Also, a mold probably would work as well, but if you want a rotationally symmetrical mold, you'll need an existing object anyway. A lathe doesn't require a round object.

[tim jones]

@SilentGuy: A taper of 1:10 would be about optimal, 1:6 would be acceptable. Any shorter and the pressure losses start to approach those of an orifice plate. As for orifice size, I'm hoping to start with something about 1/8" and drill out to find optimum range.

[Silence]

Thanks for the figures. A ratio of 1:8 would mean going from 1" to 1/8" in diameter with 3/5" of length. That's workable. You may not be able to find a female mold though; actual cones are much more common.