Apparently it is "traditional" that dragon costumes should have "flaming" breath. (If not a tradition, it is at least expected.) This is most commonly done by using a CO2 extinguisher to blast out a cold white fog. CO2 extinguishers are not cheap, and one must come up with a method of discharging it while in the costume. This means a way to carry it around, triggering, and then getting the CO2 from the extinguisher to the mouth of the costume must be devised.
While nothing can be done about the cost, the other issues can be addressed without too much difficulty. The smallest CO2 extinguishers that are available are a 5 pound or a 10 pound version. Both are designed to be hand held, the 5 pound having a rigid copper tube funnel pipe, and the 10 pound having a short reinforced rubber hose with a plastic funnel on the end. The actual weight of these is greater than 5 or 10 pounds, the weight refers to the amount of liquid CO2 that is used to charge the cylinder. Both of these extinguishers are rated to give about 5 minutes of discharge time (if used continuously.) The weight of the 10 pound extinguisher is a bit much, when added to the weight of the costume, so I focused on using the 5 pound version. It is not practical to have the extinguisher mounted high in the head of the costume, so will be mounted lower, in a backpack arrangement. This should also help offset the weight of the head and neck, acting as a counter-weight. I ordered a custom rubber hose from a local supplier, with the correct threaded connectors. A small pulley is mounted on the handle of the extinguisher. By running a cable through the pulley, and back down to the handle, enough leverage can be applied to remotely trigger the valve using a bicycle hand brake assembly.
To enhance the "flames", I added a ring of bright yellow, orange and red LEDs to the end of the extinguisher nozzle. Four sequencing circuits were modified to give different flashing rates, then the LEDs wired to the circuits. The LEDs, and circuits are hot-glued to the end of the nozzle. A mounting plate for a small lever-action micro switch was made to fit over the nozzle, then a thin brass fin soldered to the lever. The switch is then mounted with the fin inside of the nozzle. When the extinguisher discharges, the pressure lifts the fin up against the outside of the plastic nozzle, activating the LED flashers. The whole assembly was then covered with plastic tape to protect the wires and circuitry. Two "D" cells are all that is needed to power all four flasher circuits.
Before getting involved with CO2, I examined several other possibilities.
Model train smoke units: I looked into the smoke units used for model trains. In exploring this possibility, I found plans on the Internet on how one person built a smoke unit for a model paddlewheel steamship. The unit is a fairly simple construction of a thin glass tube, a fluid reservoir, and a heating element. A piece of nichrome wire coiled around the upper length of the tube heats the fluid to vaporize it, creating the smoke. Commercial smoke fluid can be used or a very light weight oil (which is what most of this type of smoke fluid appears to be.) The glass tube draws the oil up to the heating element through capillary action, although some designs may use a fiberglass wick to the same end. I built and tested a small unit, and while it worked okay, the volume of smoke is fairly small, at least for what would be desirable for a large dragon costume. There is also the concern that burning hydrocarbons are not the most desirable thing to have inside a poorly ventilated costume.
Dry ice fog: I discarded this idea without testing, although the idea of a CO2 fog "leaking" from the mouth and nose had a certain appeal. It would be subtle, and fairly visible. But it would also require some method of insulated containment in or near the head. It would require its own venting to force air through in a controlled manner, and for the best effect, would need warm or hot water dumped on the dry ice to create any volume of fog.
Ultrasonic fog: These are small ultrasonic units that vaporize water in a small local area. Again, a water reservoir would be required, and the fog volume is very small and subtle. The appearance for costume breath would be minimal.
Powder based "fog": This method calls for a small jar of baby powder or cornstarch (the latter is recommended), with a two pieces of tubing going in and out, an air hose is attached to a small hand pump or bladder. When a blast of air is forced through the hose, it is either diverted through the jar of powder, or across an opening like a spray paint gun or air-brush. This stirs up the powder and creates a visible white cloud. The disadvantage here being the build-up of powder on the costume, possible allergic reactions from the audience, and the inability to have a long continuous blast. The length of the blast could be extended by using canned air or a small CO2 tank to drive the dust, instead of a hand-driven air bulb.
Smoke machines: Most smoke or fog machines would be bulkier than a CO2 extinguisher, and they require access to 110V line power. They would require a short warm-up period, and cannot be easily started and stopped. While some use dry ice, others use smoke fluid, usually an oil-based liquid, and are typically combined with steam or a water-based fog. In addition, a small version of this type of system would cost more than a CO2 extinguisher.
Smoke in a can: This has some possibilities, although again, the effect is subtle. Smoke in a can does not create thick dense clouds of fog or smoke, but creates a hanging mist that, while effective for lighting and laser effects, would not be readily apparent as an effect.
If the mist is passed through a section of tubing, the force from the nozzle pressure is reduced, and the mist is concentrated somewhat. Although still not very visible, this yields a subtle, but visible effect for closer viewing. To remotely control the aerosol can, a plastic handle grip (Rustolium® brand) was attached. A cable runs through a hole drilled in the handle and trigger, and a crimp tube holds the cable in place. The other end of the cable is attached to a standard bicycle brake mechanism that will be mounted on a section of 3/4" copper tube. Corrugated sump pump tubing fits neatly inside the opening at the front of the handle, to feed the mist up to the nostrils. Hot glue secures the tubing in place and reduces mist leakage.
Testing shows leakage to be a definite problem. The mist is a food-grade mineral oil. The hose traps enough of the mist close to the can that the oil builds up and leaks out around the base of the handle. If used heavily, the oil would also build up and drip from the nose as well, causing discoloration and staining the fabric of the costume. I will use the hose, but will explore using a powder based "smoke" with it instead.
Super bright red, yellow and orange LEDs are hot-glued to a plastic ring that slides over the extinguisher nozzle.
The LEDs are wired to flasher circuits, modified to flash the LEDs in sequence at different rates. Connections are made "random" to enhance the flame like flickering effect.
The LED ring is hot glued to the end of the extinguisher nozzle. A brass mounting plate for the micro switch is screwed to the nozzle, and wired to power the flash circuits on when triggered.
A brass fin soldered to the lever of the switch increases the surface area, so that the switch is turned on when the CO2 is discharged.
Plastic tape covers the wires and circuits for protection during handling and mounting in the costume.
Cable ties hold the hose in place.
|Powder based "smoke"
A powder-based smoke effect can be created with a few simple items. A container to hold the powder, tubing to feed compressed air in and out of the container, and a source of air or compressed gas. A test container was made using a small plastic jar, two pieces of brass tubing, and a can of compressed air. The first thing I found was that if the input and output tubes are the same size, the pressure blew the lid off of the container. I used some 1/8" ID vinyl aquarium tubing for the test. To correct this, I used a larger output tube, but stayed with the 1/8" input tubing for the final design. A small reusable juice container holds the powder reservoir. I removed the sipper tube, and drilled a large hole, into which I placed a brass hose fitting. A soft rubber tube extends down into the container, and a section of 3/8" tubing feeds into the existing white corrugated hose in the head. For testing, I used a length of green silicon rubber tubing from the can of compressed gas to the input tube. Hot glue secures both hoses to the lid of the container. In the final assembly, rigid PVC tubing will extend up the neck to the bottle placed inside the head.
How it works: The pressurized gas blows into the container, churning up the powder, the air flow continues out of the container through the larger hose, and carries a small amount of powder with it, until it exits from the nose in a small cloud. The powder is a mix of ultra fine glitter and titanium oxide, but you could easily substitute talcum powder or corn starch. The air pressure is too low to carry much regular size glitter all of the way out of the nose, but the ultra-fine glitter works well. Glitter settles to the floor much slower then the powder, although the powder is more visible.
Besides regular size glitter, I also tested a variety of other materials to use for the smoke effect, including powdered charcoal. Any of the white powders create a much more visible cloud than the charcoal. The density of the powder also impacts on the visibility. Cornstarch is slightly less dense than either titanium oxide or talcum powder. I chose the titanium oxide for the reason that it is ground much finer than talcum powder, it is bright white, and already I have a big bag of it.
Follow-up Notes: Ignore the pressurized air can. Blowing into the tube works fine, and it is one less thing to try to position or locate when inside the dark costume.
Mounting the CO2 funnel
To extend the extinguisher output up to the mouth, I ordered a reinforced rubber hose from a local supplier. The 5 pound extinguisher uses 1/8" standard pipe threads, but the smallest hose connections I could get were 1/4". A stop at Sears hardware for various 1/4" to 1/8" bushings, elbows, and connectors, gave me all the pieces necessary to mount the funnel in the head. The resulting assembly has the rubber hose coming up the neck into a 1/4"-1/8" bushing insert, attached to a 90° 1/8" brass elbow that is connected directly to the copper tube of the funnel assembly. Before mounting, I covered the red tape with black electrical tape to provide another layer of protection to the wires and circuits. A quick-release hose attachment was added to the extinguisher, allowing the hose to be snapped to the extinguisher quickly, and without having to untwist the hose with screw-on connectors. One side effect of adding the length of hose, is that pressure builds up in the hose and then releases slowly. With the funnel attached directly to the extinguisher, the CO2 blasts start and stop abruptly as the handle is pressed.
The completed powder "smoke" assembly. The output of the compressed gas connects to aquarium tubing to a reusable juice container. Larger tubing exits from the container through the larger tubing to the nose. A piece of foam seals the gap between the 3/8" tubing and the white hose leading the nose. A mixture of ultra-fine glitter and white titanium oxide powder is in the container.
Quick disconnect assemblies were added to the hose and extinguisher to simplify hook-up inside the costume. The control cable feeds through a hole drilled in the lower (fixed) handle, through a small pulley attached to the upper handle then back through a second hole in the lower, and fastened with a crimp connector. The leverage created by passing through the pulley makes it easier to fire the extinguisher.
Follow-up notes: The cable stretched on the first use. I had to reach behind me to fire the extinguisher manually. I removed the pulley and ran the cable through the handle directly for the Conglomeration masquerade. This time the cable end popped-loose from the bicycle brake handle. Again I had to reach behind and manually fire the extinguisher.
[Back to Valhalla] [Dragon Index]
Web site images and content is Copyright © 2000 Chris Kramer unless otherwise noted. All rights reserved