There are many alcohol stove designs, including some that use Altoids cans, but I haven't seen another one like this. I've been using it for a few years and have been happy with it. The thing I like best about it is that it is really easy to make. So I thought I would share it.
Cut a cylinder out of the beer can and another cylinder out of the V-8 can. The cylinders should be at most 2.0 cm (0.75 in) tall. When set in the Altoids can, they should almost reach the rim. Good instructions for the technique of cutting clean cylinders out of aluminum cans are in the instructions for the Pepsi can stove and others.
|If you don't have a V-8 can, cut a
second cylinder from a beer can and then cut the cylinder
to make a strip 17-18 cm (6.25-6.75 in) long. Cut slits
about 0.5 cm (0.25 in) from each end of the strip as
shown here. Each slit should be half the width of the
strip. Slide the two slits into each other to form the
second cylinder. It doesn't need to be glued.
Cut deep notches into both cylinders like this:
The notches should have flat tops, not pointed, so they aren't a hazard. Put the notched cylinders in the Altoids can, pointy side up. Don't glue anything. The stove is done. It weighs 32 g (1.1 oz).
|Explanation: The aluminum cylinders transfer heat from the flame to the alcohol, which makes the alcohol evaporate faster, which makes it burn faster, which makes it heat your food faster. The notches allow air to flow in and mix with the alcohol vapors--alcohol needs air to burn. I experimented with more or fewer cylinders, bigger or smaller cylinders, notches, no notches, wicks, etc. and this worked best.|
You can make a "simmer ring" (it's not really a ring) from aluminium flashing. Place an altoids can on a piece of aluminum flashing, letting the can hang over the edge of the flashing about 1 cm (0.5 in). Trace around the edge of the can with a felt-tip pen. Cut a "C" shape about 0.5 cm (0.25 in) outside of the line you drew. Round the corners. Smooth the edges by laying it flat and sanding it just enough to knock off the rough edges. If you sand it too much you will give the aluminum a sharp knife-like edge. The simmer thing weighs 3.2 g (0.1 oz).
Operation: Pour some alcohol into the Altoids can, light it, and start cooking. No warmup is needed. Use the pot, pot stand and windscreen of your choice. I recommend using matches to light it because I always singe the hair off of my fingers when I use a lighter. (Yeah, I'm a Neanderthal.) To use the simmer thing, just set it on the stove after the water has started boiling. Adjust the heat by sliding the simmer thing left or right. You can "turn off" the stove by removing the pot stand and windscreen and setting the pot or its lid directly on the stove. If this doesn't snuff out the stove then either the aluminum cylinders are too tall (so trim them a bit) or your pot or lid is warped (so use something else). Put the Altoids lid on the stove before packing it to keep the cylinders from spilling out.
WARNING: Alcohol burns with a nearly colorless flame so you occasionally might not notice that it is still burning, especially when the alcohol is almost used up. If you try to pour unused alcohol back into your fuel bottle or pour more fuel into the stove while the alcohol in your stove is burning, or even while the stove is still hot, fumes in the fuel bottle might EXPLODE and splash burning fuel all over you. This would be BAD. It would HURT and RUIN YOUR TRIP. This type of accident HAS HAPPENED in high school chemistry labs, resulting in serious injuries and a few deaths. Also, if you use ultralight silnylon gear, bear in mind that silnylon is very flammable. Remember: stop, drop and roll.
I have used this stove for about 500 trail miles. The only problem I had was when the Altoids can had corroded in storage so it was hard to open. That's what I get for not testing things before hitting the trail. Now I get a new Altoids can from time to time. I am still using the first set of aluminum cylinders I made.
This stove boils 1 pt (473 mL) of water in just under 5 minutes using a Wal-Mart grease pot with lid and a combo windscreen/pot stand that is 6.4 cm (2.5 in) tall and 16 cm (6.25 in) in diameter. It is hard to compare alcohol stoves because the boil time test is affected by pot stand height and windscreen design (see below), as well as altitude, the initial temperature of the water, the wind, the type of denatured alcohol used (there are many), how accurately the tester measured the volume of water, etc. The Altoids stove isn't the fastest or lightest stove, but it heats quickly, it's light, rugged, compact, and it sure is simple. If you had to, you could make one at a resupply stop--it would give you a reason to buy candy and beer. Just make sure your Swiss army knife has a scissors to cut the aluminum cylinders.
With my pot, the stove gives its greatest heating rate with a pot stand that is about 8-9 cm tall. At lower heights, it heats more slowly but also uses less fuel so gram weenies like me might want to use shorter pot stands. Unfortunately, the design of the windscreen affects this so your mileage may vary. These data were gathered using a typical windscreen: aluminum foil with a row of holes around the bottom and a 1-cm gap at the top. Here is a spiffy graph.
This graph shows three things. First, a short pot stand transfers more of the flame's heat to the water because the pot is closer to the flame. That is why fuel efficiency is greater for shorter pot stands. Second, a short pot stand makes the fuel burn slowly by restricting air flow. That's why the fuel consumption rate goes up as the pot stand gets taller. Fuel efficiency goes down, though, because the flame is more spread out and more heat escapes. Third, when the pot stand is too tall, so much heat escapes that the alcohol in the stove is not heated as much, which means that it doesn't evaporate as fast and doesn't burn as fast. That is why fuel consumption rate goes down for the tallest pot stands.
I came up with this stove after spending a couple of weeks working on stoves in my spare time when I should have been doing something more productive. (Playing with gear is not as good as hiking but it sure beats working.) I teach chemistry at a college so I had good tools. I would put the stove on an electronic balance to record the weight change of the fuel as it burned and put an electronic thermometer in the pot to simultaneously record water temperature. From this I could calculate water heating rate, fuel consumption rate and fuel efficiency. For hungry, impatient hikers heating rate is the bottom line. For gram weenies, fuel efficiency is the bottom line.
I experimented with several stove types including the Photon stove, Pepsi can stove, V-8 stove and a bunch of things derived from the Sgt Rock cat stove and hobo stove before settling on this design. I found that elaborate designs don't do much. An alcohol stove is not rocket science. It evaporates the alcohol and mixes the vapors with air so they can burn. (It's the vapors that burn, not the liquid.) Basic thermochemistry says that the chemical energy in the alcohol is the same for every type of stove. Some stoves burn the alcohol faster, though, so they heat your food faster. The best stoves were good at transferring heat from the flame to the liquid alcohol to make it evaporate faster and burn faster.
The key to this stove is the aluminum cylinders because they conduct heat from the flame to the fuel. I experimented with the number and size of the cylinders and the presence or absence of notches. I used aluminum because it is cheap, light, and has good thermal conductivity. I tried copper foil, too (copper has greater thermal conductivity than aluminum) but it didn't make the stove any better. I also tried different pot stand heights and played with the design of the windscreen. All of these things interact with each other. I did not come up with a global optimum design of everything (answer = 42) because I got sick of the project. Here are some of the factors that affect an alcohol stove setup, and a short list of the things one might want to optimize.
If anyone wants a project in statistical experimental design to optimize a system with six variables for four responses, here it is. Warning: I figure you should try at least 56 (footnote*) separate setups using different combinations of the factors. See why I gave up? Even playing with hiking gear isn't that interesting.
Here, for your amusement, is some info about the energy content of popular fuels.
|propane||46.2||23-33||(depends on pressure)|
|white gas (gasoline, petrol)||43.6||32.1|
|ethanol||26.8||21.2||(watch out for water)|
|hexamine (Esbit, etc.)||30.0||n/a|
"kJ" is a measure of energy. kJ/g is the energy/weight density of fuels. kJ/mL is the energy/volume density of fuels. These numbers show that propane, kerosene and white gas, which are all hydrocarbons, pack the most energy per gram and the most energy per liter. However, the equipment you need to burn them is heavier, so you end up carrying more weight except on long trips. I did some calculations comparing ethanol to white gas and found that if you have to boil more than 30-70 pints of water (depending on the weight and efficiency of the stoves) alcohol fuel + equipment weighs more than white gas fuel + equipment at the start of the trip. For me that's a two-week trip without resupply. As time goes by, you burn fuel so your load gets lighter. Esbit is less efficient than the hydrocarbons but more efficient than alcohols. The main advantage of Esbit is that the stove can be really simple and light. Here are some sites that do a very good job of comparing the weight of different stoves and their fuel on long trips: Stoveweight vs time over 14 days, Stoveweight vs time over 28 days.
Among alcohols, ethyl is a lot better than methyl because it's much more efficient and also less toxic. Isopropyl alcohol really isn't practical because it is less flammable, more toxic, and more stinky. Also, the isopropyl alcohol you buy in stores usually has a lot (30% to 50%) of water in it. Water in the fuel is bad because (a) it doesn't burn (b) it's heavy and (c) it steals heat from the flame. (Ask a chem major about enthalpy of vaporization.) The "grain alcohol" you can buy in liquor stores is mostly ethanol but has about 5% water. "151" rum is 25% water so it's even worse. Denatured alcohol is the way to go because most types have less than 1% water. There are a lot of different types of denatured alcohol but all are basically ethanol with stuff added to make it poisonous so don't drink it! The denatured alcohol I have found in stores is a mixture of about 95% ethanol and 5% methanol.
I also used data from REI to estimate the efficiency of the MSR Simmerlite, Whisperlite and XGK white gas stoves. The Simmerlite and XGK were about the same. The Simmerlite is less fuel efficient but lighter. The XGK is heavier but more fuel efficient so it's better for long trips. The Whisperlite combines the disadvantages of both--it is heavier than the Simmerlite but less fuel efficient than the XGK. None of these is as light as an alcohol, canister or Esbit stove, though, except maybe on very long trips without resupply stops. If I were planning to through-hike the Yukon, I would probably get an XGK and burn kerosene. White gas or kerosene stoves are preferred for cold-weather camping and mountain climbing, too. For most hikes, though, the light choice is between alcohol and Esbit.
And another thing against white gas stoves: they are fire hazards. Flares are expected, but sometimes the flares are spectacular. I once saw a whisperlite or simmerlite produce a three-foot (1 m) fireball. It didn't last long, but if the wind had been blowing the other way the guy would have been in trouble.
* This is for true geeks. The science of optimizing a system is well developed, whether the system is a chemical reactor or an alcohol stove + pot stand + wind screen. Instead of trying every possible combination of factors, you try enough of them to derive an equation that describes how the factors interact with each other and then use the equation to find the best design. A second order equation with first order terms for factor interactions describing a six-dimensional response surface has 49 parameters (2*6 + 1 + 62) so the absolute minimum number of experiments would be 49. You should have at least seven extra experiments to let you test the statistical significance of your model with some confidence, making it 56 experiments. A really thorough star + factorial design would have 77 experiments (2*6 + 1 + 26). Anything less might not find the best design. If you enjoy doing this so much you want to do it again and again, the optimmum design is probably different for different stoves. Alternatively, you could do a systematic search guided by a simplex or steepest-descent algorithm, in which case you would have to build a new custom setup specified by the algorithm before each experiment. That might be quicker. Or it might not. Personally, I think that "good enough" is just fine.