The Forge Burner Project

Building a forge is fun but difficult, at least from my experience. There's burner pages online that give one the impression that these things are easy... Sure. For example: This one or this one or these. This page is not a series of success stories, with the failures left out to protect the dignity of the author, but instead, a raw exposition of just how difficult these things can be.

The problem I have is: When the forge gas exit is closed off a bit to better contain the heat inside, the pressure within the forge increases, seriously affecting the venturi style burners' ability to draw in sufficient air. The mixture gets too rich and flames start shooting out of the forge at every crack between the firebricks, and a huge flame at the exit! Also, the burner works great when running in open air, but gets rich when shooting into the forge, I suppose from the forge exit restriction and resulting backpressure.

There must be a way to either understand how venturi burners work to make them work better, or design a fan assisted burner so that forge loading and exit port dimensions don't affect the heat input or reducing/oxidizing conditions.

If you try working with these things, you will find that they are finicky. You feel like you need to know all kinds of things that will allow you to actually engineer them, after having a few successes and many more perplexing failures. After all, if it isn't working right, it wasn't engineered properly, right? From what I've found, these are the relevant values:

Propane/air flame velocity = 16"/second, or 0.4m/s

Gas to air ratio for complete combustion = 1:25 by volume

Gas requirement for 1,000 BTU/hr = 3ml/s (at atmospheric pressure)

Propane delivery through 1mm orifice at 10PSI ~ 300ml/s

Air requirement for 1,000 BTU/hr burner = 0.17cu ft/min

Gas delivery is proportional to the orifice area.

Gas delivery is proportional the the square root of pressure.

Propane/air autoignition temperature = 580C


The burner is supposed to have a flare, which proves to be true in my experimentation, supposedly so that the gas mixture delivery at the exit of the burner is at the flame propagation velocity of 16" per second. This doesn't seem to be true for my designs.

As I understand it, gasses burn by deflagration or detonation. The former burns at the flame front velocity of 16"/s (for propane/air), but detonation means a flame front traveling at the speed of sound! Since my burners are noisy, I'll expect that they are not burning nicely like a flame at the stove, but occasionally, randomly, and noisily 'jumping' toward the mixing tube, then stopping because... Well, I haven't figured that out yet, although I suspect an acoustic mechanism where the local pressures fluctuate due to the 'jumping', causing the detonation to stop. The flame is noisy, which isn't the case at my stove, and the gasses coming out of my little burners is definitely raging at more than 16" per second before being lit. I've built burners that 'puff' a few times a second, ones that 'screech', and many that unexpectedly 'blow out', so I'm of the mind that there's something really complicated going on.

I expect this to be a mechanism that is so complex that engineering by the numbers is the wrong approach! We need to experiment.

I tried shooting propane down various tubes with various orifice sizes, but the burner that I like best is a 1mm gas orifice placed inline within a 1.5" diameter chunk of conduit:

It looks like this from the front and back, and then there's the venturi I tried to put inside:

The front end is two pieces of 1/16" steel that seems to guide the flame better than a flare; they extend into the tube by about 3/4". The orifice is on the end of a chunk of 1/8" pipe, about 2" into the outer tube, held in place with some crappy welds. As for the venturi, machined from a chunk of aluminum rod, it didn't seem to help in any way, no mater where it was positioned within the tube. -This is probably because the venturi is the wrong size, but how many does one have to machine to get it right?

Then there are the other designs I've tried:


They all work, but not well, despite screwing around with sheet metal dampers to cover the air holes.

As ugly as it is, the first burner works best, modeled after a weed burner, just scaled down. I was really impressed by the weed burner, such a simple construction, and works so well.

And then there's the silliness I've been doing lately:

The mixing tube must be too short, because this really needs tuning! The flame happens within the expansion fitting on the far right, despite the fact that I machined it out to a 10 degree taper, and easily blows out if not burning deep within the flare. No combination of air or gas control makes for a useful flame. This is really frustrating!

The main problem was that the design modeled after a weed burner would burn beautifully on the bench, but not so in the forge. When I would pile up a bunch of firebricks to narrow the forge's gas exit, the mix would go lean and flames would leap out of the exit. When I would gently blow air from my shop compressor into the back end of the burner, the forge would get hot as H*** and be basically doing what I wanted, with no external flames. THAT led to the monstrosity above, which ain't doin' it either. -Yet.

Aha! Narrowing the mixing tube size to 1/2" causes the mixture velocity to be great enough to prohibit flames from backing down the tube, and a simple flare (1/2" to 1" coupling) keeps the flame from blowing out. The blower has a pair of aluminum plates with 1.5" holes punched in them that can slide across the blower exit to adjust the air supply. The propane can be adjusted at the regulator. The combination can be run over a reasonable range of power. It is clear that the small mixing tube and the orifice size is limiting the maximum power, which gives the impression that all we need is a set of orifices and mixing tubes to get whatever power is needed. A good length of tubing from the blower to the mixing/gas/injection area also may be smoothing the airflow prior to mixing.

And then I tried it with my side-arm injector from the first blower experiment:

The flame isn't going straight in the last one, so perhaps some sheet metal on edge within the mixing tube and maybe just after the blower would help 'straighten' things out.

From my present point of view, despite the rare possibility that a venturi burner could be made that allows in-situ adjustment for forge backpressure, this blower-assisted setup is great! Yes, it needs power for the blower, but this allows for adjustment of airflow without being finicky... The only problem is that the orifice and mixing tube need to be sized for the job, but given a long enough mixing tube, and maybe some 90 degree bends that could assist in the mixing process, a compact, very powerful energy source is at hand!

It lights up the forge beautifully. If flames start to pour out of the exit, I just crank up the air a bit (or lower the gas pressure), and they come right back in again. Perfect!

And it's not so expensive. I made a few trips to the hardware store for fittings, spending maybe $10 on each trip, and the blower was about 60 bucks... Altogether the setup will have about $100 of parts (not including propane tank and regulator). The final assembly will go into a small hand "truck" I built, with the blower and a tank of propane in the base, with the pipe and fittings running horizontal from below, going up and around 180 degrees, and right into the firebrick pile on top. Very compact and portable.

Here's the finished Porta-Forge, with only a few firebricks in place:

Well, it does need power, so its not completely portable, but it is ready for a pile of firebrick, so I can heat whatever I like. Some details of the construction:

The left photo is the nozzle, which is actually machined to a 10 degree taper, but so badly that it looks like screw threads. It's a 1" to 1/2" pipe bell reducer, machined at the 1" end. The right photo is a close up of the spacers that attach the fan to a 1 1/2" pipe flange, with steps cut in the aluminum spacers to allow for the two sheets of aluminum to close off the air passage.

This has been quite fun, but it took me two or three weeks of struggling to get a grip on it, and I still haven't figured out how to do it with a venturi. Since the professionals usually use blowers, I'll expect that for my applications, a venturi burner just won't work well enough for me to even try.


Please send any questions or comments here.