Testing, tuning and using the Trianglya


Testing, tuning and theory

When I had completed the trianglya as detailed, I had a good read of Anthony Dunk's site (see links). He has been water-boil testing his Trangia burner housings, and I thought I had better do the same.

Initial testing against the classic Trangia housing showed my housing to have a boil time over 12 minutes, and a meths consumption for the boil of about 30 grams. This compared to the classic Trangia time as previously quoted, of 8 minutes 30 seconds, and a consumption of 16 grams. The initial reaction was, of course, to ask myself why I had bothered to build anything. Some tuning of the design was in order. When tuning was over, the boil time was down to less than 7 minutes, and the consumption was down to between 14 grams and 16 grams. The testing was done in still air, where a fully-shielded housing has fewer advantages

Plainly, tuning made the difference between success and failure in this case. I appreciate design theory has always been unpopular, especially with the Thermarest Generation, but I will have to put down a summary of how, and why, the classic Trangia works so well. Once that is established, the lessons can be applied to my design, or any other. Skip or skim as much as you like.

I can only cover Trangia-type burners. Jet-type burners such as the MSF run according to different rules. If you place a full Trangia burner naked on a heat resistant table and light it up, the meths ignites, and it burns very slowly of course. As the small flame starts to warm up the burner, the heat vaporises more meths, and the flame gets bigger. The flame size only stabilises when the table and the air are conducting away as much heat as the burner is giving out. Note that since meths boils at 80 degrees C, the burner remains at that temperature, but a bigger flame can boil away more meths, since it is providing more heat. We could call the process of heating-up 'thermal takeoff', since it has much in common with a 'plane leaving the ground. A similar device - the 'hot pot', is used in oil-burning household heating furnaces, but here a variable drip feed is used to control the burning rate. A 'hot pot' can take 30 minutes to reach running temperature.

If the Trangia burner were placed on a table made of thick copper or aluminium sheet, and lit, there would be no thermal takeoff, and the flame would stagger along not much above candle-size. In this case the cold table is removing virtually all the heat, so the burner never starts. If you look at the classic Trangia housing, the burner sits in the centre of aluminium sheet, but as much of the sheet is drilled with holes as mechanically possible. This greatly reduces heat flowing either in or out of the burner from the rest of the aluminium body. The burner is insulated, or 'thermally decoupled' from the rest of the body. Incidentally the holes are dual-function, since they also let in the air for combustion.

To put this into one sentence, the Trangia burner is thermally decoupled from the rest of the body, to achieve a quick thermal takeoff.

A second factor in thermal takeoff is a big cooking-pot perhaps full of ice-cold water. If that can remove heat from a burner trying to start, takeoff will slow to a crawl. The Trangia locates the cooking-pot on three steel clips, so it can be heated by flame, but minimal heat can travel by conduction through the clips from the housing. Thus the housing stays warm when burning starts, and takeoff is facilitated.

Once the burner is flaring away, the Trangia housing heats up, and of course the cooking-pot does so too. Once both of these are over 80 degrees C, heat will try to flow back into the burner, and make it run even faster. The thermal decoupling of the burner is very effective, since heat finds it hard to get through the holey area, and cool inlet air is flowing through the holes to further reduce the warming. The Trangia design thus prevents 'thermal runaway' and keeps its flame under control.

To summarise, the classic Trangia achieves a fast thermal takeoff, but avoids thermal runaway, by thermally decoupling the burner from the housing, and also the housing from the cooking-pot. It is a very good design.

Let us go back to the trianglya as first made. During the boil test it was starting very slowly indeed, so takeoff was poor. When it finally did manage to get going and the water and housing got hot, runaway was very plain, with flame spraying all round the pot. It thus took forever to boil, and wasted fuel like mad in the last few minutes.

Looking at the trianglya, the burner edge sits in contact with the aluminium housing over a wide area. There is no decoupling between the burner and the housing. The pot sits straight on the folded top edges of the housing. Unlike the the square predecessor to the trianglya, there is no decoupling between the cooking-pot and the housing. Once this was established, the answer to the problem was pretty obvious.

The burner sits in a hole made of three segments. Using pliers, I bent up one end of each segment, outlined in red on the photo, so the burner rim rested on the three bent tips about a millimetre above the hole surround, rather than making thermal contact with it. Not only did this thermally decouple the burner from the housing, but the amount of decoupling could be usefully varied by changing how much the tips were bent.

The red spots on the photo mark six 3 millimetre diameter holes. Three 20 millimetre long M3 screws, with two nuts screwed up to each screwhead were dropped through the holes closest to the burner, to make pot spacers, so the cooking-pot rested on the screwheads 5 millimetres above the housing. The cooking-pot was now decoupled from the housing. The second set of holes is for large frying-pans etc.

I ran the water-boiling tests again, repeating the test several times. The trianglya now boiled in less than 7 minutes repeatably, and used between 14 and 16 grams of meths. Both figures are a little better than the classic Trangia. Varying the bent support tips to alter the burner decoupling resulted in corresponding changes in the boil time and final flame size.

The boiling tests were conducted in still air at about 20 deg C on the dry concrete floor of a garage. The burner was filled to a total weight of 130 grams, and weighed again after the boil. The cooking- pot was filled with 600 milliliters of water. The water came from a large watering-can that was filled, and then the contents used for all tests.

Cooker insulation and 'snow feet'

During an early experiment, I was boiling water, expecting a boil time of about ten minutes. The water finally staggered to the boil in twenty minutes. I found this was due to the burner legs, which were standing on the concrete floor as usual, but there was a small amount of water between each leg and the concrete. The water enabled fast heat transfer between the housing and the concrete. The housing stayed cooler than usual, so the burner never took off. Although the decoupled burner will be less sensitive to housing changes, this does emphasize the cooker will be sensitive to what it stands on.

If your cooker stands on rock, dry vegetation or sand, and conditions are dry, heat will not flow out of the legs, and it will behave as it should. If the ground is wet, marshy or of course snow-covered, you will have to insulate the legs. Jamming a short length of stick into the hollow legs, so the leg ends are clear of the ground, and the cooker rests on the ends of the sticks, is a quick fun project for that underused Swiss Army knife with the dinky little saw. Snow feet for use on soft cold snow protrude at least six inches from the end of each leg. The cooker is wiggled down on the legs to the right height in the snow, which is then tamped-down around the snow feet. Normally cookers melt into the snow with amazing speed. A trianglya on snow feet just sits there without moving while you cook.


Still air, so common in garages during test boils, is rare on camping expeditions. The burner, once it has taken off, will not blow out easily, but you can expect your meths bills and boil times to rise sharply. For really windy conditions, the classic Trangia is still the best tool for the job. Light gear always demands more skill and more effort, which is how you pay for that easier packload. Follow the book advice to shelter your cooker, without taking it into the tent. Some of us carry a trowel, so we can leave the Woods as we found them. A hole around nine inches diameter and ten inches deep will take the cooker comfortably, and create that precious block of still air. If you are on snow such a hole is very easy. I've dug a whole tent in to roof level in a snowbank, to get out of an astonishingly strong wind, but that did take an extra couple of hours.

Melting snow

Melting snow is not in the same department as cooking. I potter around in the dark with my burrlight on my hat after dinner, with plenty of time to spare before bed. Often the wind has died as well. What matters is not how fast you can melt snow, but how much you can melt for every gram of meths. I filter the melted water rather than boil it, so the cooking-pot never goes above zero deg C all the time melting is going on. You need, as always, still air, and a low flame. If you are getting value for money, you will not be able to feel any heat rising past the cooking-pot at all - the melting snow just swallows every calorie. I suspect this operation will go best with the pot spacers removed, which is one of the reasons I made them drop-in. This thermally couples the cooking-pot to the housing, so the housing never warms up. No heat is lost into the air, unless it is way below freezing. The burner remains decoupled from the cold housing, so it can warm enough to take off to some degree. In this case the restrictor would be fitted. I'll work out the details in the coming winter.

The snuffer

The exact cat-food can used holds feline snacks from 'Fancy Feast' and is labelled 'product of USA'. If the can is carefully cut down to 22 millimetres depth with tinsnips or an unloved pair of scissors, it takes up minimal space, and is a perfect push fit over the rim of the restrictor. The cut-down can snuffs better, but operation is still not perfect. It takes skill to drop the snuffer square onto a flaring burner, each and every time. A clean drop will immediately snuff the flame, but if it lands crooked, by the time you have pushed it straight with a piece of stick the snuffer has got hot, and the flame will not go out. A splash of water or small lump of snow dropped on the top of the snuffer will put out the flame. This provides a short- lived cold area where the meths vapour under the snuffer condenses. This is enough to immediately extinguish the flame. Since the lip of the burner is raised under the snuffer, no water gets into the remaining meths.

Practical trials

Inevitably, a few camping trips show up design aspects that were not evident on the garage floor. You can expect this section to change over time. Warming wash water on a mountainside created no major problems, and meths use was minimal with the restrictor fitted. The tripod stood anywhere it was put, but if the cooker was not level, a 1 litre capacity thin stainless-steel saucepan with a high-gloss coppered base had a tendency to slip sideways off the top support pegs. Resetting the stand so it was level was one answer, but just as easy was to stand the edge of the pan on the lowest side just inside a support peg, and resting on the top edge of the stand. This peg stopped the pan from slipping off the edge, and the pan itself being off level had no noticeable effect on water heating. The first fumble and fudge test has been passed.

Afterthought - Lighting Trangia cookers

The Trangia meths burner has no moving parts, but has its own peculiarities when it comes to lighting. Under normal still conditions a match will suffice, and there is a gentle 'thud' as the invisible flame starts up. When the temperature drops below freezing you need to ignite something like a dry leaf or grass stem dipped in meths before dipping it gently into the meths in the burner, or holding it just above the meths surface.

Although I carry a small sealed tin of firework-type matches from the boating shop for dire emergencies (they have never been opened) even camping matches are far from ideal, and I have given up carrying them. I carried a small BIC disposable butane lighter for years, before buying one in Vancouver with a dodgy valve that only just lit, and would permit no adjustment. Not a product to trust your life to.

The Zippo petrol lighter is an American icon that is still being made. The big solid flint wheel and wick plainly will run for ever, but if you try filling the fuel wad with lighter fluid, you find the lighter will light for about three days before it has all evaporated into our suffering atmosphere. No wonder butane lighters took over.

After much thought, I carry the Zippo dry, which is no danger to anyone. When the Trangia burner is set up and filled with meths, I dip a stick into the meths, then touch the end of the stick to the Zippo wick, which transfers a couple of drops across. The lighter flint wheel then ignites the wick, which ignites the flaming end of the stick, which in turn will light the burner under any conditions.

So far I have not faulted this ignition arrangement. You are welcome to copy if you like this way of doing things, or have a good laugh at my expense if you don't.