Gas Forge Build 1: Concepts and Burner Tube

As a knifemaker progresses, naturally we want to explore new areas. I have an interest in making  pattern welded steel. In order to do this one needs hot and fast heat. The electric heat treatment oven is not quick enough to do this, so a gas forge is the answer.

WARNING: I am offering this information for entertainment purposes only. I will not be liable for any use or misuse of this information. This project deals with multiple hazards including flammable gas and substances that you must understand and work with at your own risk. Please consult a qualified gas fitter when making any connections to LPG or other fuel systems.

A typical DIY gas forge is consists of two major parts, the forge body which has some insulation and openings for putting metal and fire into it, and the burner(s) which create the heat to make things happen.

The forge body is constructed either with a steel cylinder lined with ceramic wool and coated with a protective refractory cement or a steel frame holding insulated fire bricks.

The general design you choose may have more to do with available materials. I have scrounged a 5 gallon air compressor (see Forge Body) and landed some free ceramic blanket from my local refractory wholesaler, so my build will be a cylinder shape.

The forge body cylinder can be made from old Freon tank or a 20 pound propane tank. Using these cylinders have their hazards and need to be treated with care before and during any work on the tank. Do your research and understand the hazards.

I decided that I don't need a giant forge for knifemaking. Perhaps 16" long by 5" x 5" square or about 400 cubic inches. A typical efficiency 3/4" burner will heat about 300 cubic inches, so I am going with two burners. This will help even out the heat along the forge and it will allow me to block off the back side of the forge for single burner half forge. I also want to be able to use the forge upright for melting aluminum and other alloys. 

Now on to the burners...

Venturi Burner

A Venturi burner consists of two main parts: The burner tube and the gas accelerator or what some call the torch.

The burner tube is the outer part of the burner where the LP gas is sprayed into causing air to be drawn in and mixed with the fuel. The end of the burner has a wider opening called a flare or a nozzle to encourage the air intake and stabilizes the burning.

The torch is the part where the LPG is jetted from. The gas is under some low pressure (usually less than 20 psi) and comes out a very small hole, typically 0.025" to 0.060".

A couple of popular designs are shown below. (Note that some details have been left so we can better see how these burners work.)

The Sidearm burner uses a reducing tee where gas is injected into the reducer and air is drawn in (entrained) from the side opening of the tee. An air choke can be made by covering the side opening with a metal disc held on with a screw.

Ron Reil's original burner uses a capped pipe across the bell reducer with a very small hole drilled in it. The LPG is jetted out of the tiny hole and down the tube. Air is sucked in from behind the cross pipe and mixes with the fuel in the tube where it comes out burning at the end of the tube..

Both of these designs are very popular for the DIYer as the parts are easy to acquire and they are proven to make an effective burner.

Some interesting things about LPG combustion to note.

  • LPG is a mix of Propane and Butane.
  • LPG requires 5 times its volume of oxygen burn correctly.
  • Atmospheric air contains about 20% oxygen.
  • Thus, LPG requires nearly 30 times its volume of air to burn correctly. 
  • The liquid in a tank boils and the vapour is burned
  • The rate of boiling is dependent on the tank area and ambient temperature.

You can see that 100 times the air is needed and this is why most burners have large intakes for air. This can be reduced by adding air via a blower or compressor, but that requires extra equipment and increases the complexity.

Burner Tube

My burner is not quite like the Sidearm or Reil design. It is more like Michael Porter's design meets Ron Reil's design. Note: For most of the images I am showing making only one burner tube. I am in fact making two.

On order to ensure proper alignment of the gas delivery tube inside the I made a jig. I cut a plug from a 2x4 with a 1-3/4" hole saw.
Here is the plug. After removing from the hole saw, I drilled the hole out a little more to 27/64". A length of 1/8" NPT nipple should fit snugly in this hole.
I cut a 1-1/2" x 3" nipple in half. Shown are the two pieces roughly 1-1/2" long. I deburred the cut ends and sanded inside to make it smoother.
For the clamping tube I marked along a some 3/8" schedule 40 pipe. The marks are 1/2" apart. I drilled and tapped the first and third mark for 10-32 screw.. The setscrews I have are 10-32. These will be fine for holding the torch in position.
 Holes are tapped. Now to cut on the second and fourth mark to make two 1" sections. Deburred of course.
This is my layout for the bracket. It has a small notch to help position it for welding. I have traced it on a piece of  1/8" steel.

This is an approximation of the torch bracket made from 1/8" mild steel.

 Cutting out the bracket on the band saw.
Shaped the curves a little better on the grinder and tweaked the notch with a file. 

Using a cheap 27/64" drill bit, the wooden plug and a nut, the tube is aligned and readied for welding. If the plug is not fitting perfectly in the centre of the nipple, some 0.02" shim stock or wraps of foil will help hold it in place.

After welding, you can tidy the surfaces up with a disc grinder. 

Now the fun part. Like porting and polishing an engine head, make the surfaces smooth and try increase air flow wherever possible. I used a die grinder and various abrasives bits to shape the bracket and round the clamping tube.

Shaped a bit better to improve airflow.

This is the view inside the nipple showing the bracket and clamp
For the flare I have some 1" schedule 40 stainless pipe (Metal Supermarket) that slips perfectly over a cleaned 3/4" pipe.
Drill and tap to 10-32 at 3/4" from the end, 120° apart around the circumference.This is so we can slide the flare on the nipple and lock it in place when were tuning. You could probably get away with two set screws at 180°.

Cut a 3/4" x 18" black iron nipple in half. Each piece will be very close to 9". Deburr each end with and clean the hole with a round file. I sanded up mine for some shine, but that's optional.

Photo is showing 1/2 of  the 18" nipple.
Screw the reducing adapter on to the threads.

Looking down inside the 3/4" to 1-1/2" reducing adapter. Check for burrs that will disrupt the airflow and file or grind them smooth where needed. Dremel works well in here.

This is the burner tube ready to install the gas accelerator/torch in.

Next we'll look at the gas delivery system.


Unknown said...

Do you need to flare out or taper the 1" schedule 40 pipe at all or do you leave it straight?

D. Comeau said...

Ideally it will flare. Mine is shaped inside the tube, then the flare slips over it so it emulates a flare. If I recall correctly the believe the slope is supposed to be 1 over 12. This creates sufficient vacuum to sustain the venturi over a wide range of pressure.

TrentO said...

I built a pair of burners based on your design. I made the mistake of porting the flare end of the 3/4" tubing and the 1" stainless. The end result was unable to sustain a good burn and had the flame front sitting just inside the 3/4 tubing. I did an experiment running the burner flare tube backwards ( 3/4" to 1 1/2" reducer as the flare end and it had a great flame. I had to go back and square off the ends of the 3/4 and port the 1" stainless quite a bit to get a consistent good flame. My plan is to flare out the stainless further with some hammerwork once my anvil arrives. I think the hard transition from the 3/4" to the 1" stainless creates a turbulent front which stops flame propogation and keeps a good flame front.

Unknown said...

"For the clamping tube I marked along a some 3/8" schedule 40 pipe" is the clamping tube really 3/8" it looks bigger in the picture.

D. Comeau said...

Yes, schedule 40 3/8" pipe is close to 1/2" inside diameter. Not to be confused with 3/8" tubing which is 3/8" outside diameter. There is a good explanation here Best wishes,


Jerry Klages said...


Totally appreciate the scope and depth of information you have offered to share. I'm looking at setting up a forge in my shop in the next year. Do you know of any relatively easy, reliable designs/plans for a forge burner using natural gas? My shop heating will be natural gas so I thought about the practical concept of using natural gas as a forge fuel.


Unknown said...

Being a bit picky about your maths bur other than that i like the whole article.

5 x 20% gives you a 25:1 ratio for a stoichiometric ratio but an extra 10% or so is recommended so you're aiming for a 30:1 ratio

20% = 1/5
5 ÷ 1/5 = 1/25

Unknown said...

5 DIVIDED BY 1/5 = 1/25???

Unknown said...

I built a forge like this thanks Dan,
Great design and great website
I used what I had here lying around and ended up with 1/2 copper and compression fittings and instead of welding.tip I drilled a 0.1 mm hole in copper but apart from that all like yours,works great
I seem to have a small cooler spot on big blades I try heat evenly all at once like for heat treat and if I was to do it again,and I will,I'd angle the burners slightly to create a swirling vortex type flame,maybe even angled towards the back and vented as those burners are powerful wouldn't want them even slightly pointed near face
But thank you great job

Lori said...

Dan, thank you for your blog. It's nice to see people close to home doing great things.

I'm using your design to build my first gas forge. The only question I've got (so far) is how you honed out the stainless piece to put a flare inside. I've gone through a couple of dremel grinding wheels and I've got a bit of a flare, but not much.

Thank you,
(also in Edmonton area).

D. Comeau said...

Hi Gordon,

It doesn't take much of a inside taper on the burner tube. The fact that the flare is larger than the burner tube gives it a head start. I used a half round file and worked around the inside maybe a dozen times. Really more to smooth the transition. Maybe not 100% efficient but pretty effective for the effort.