Recommendation? Thermocouple in my forge for heat treatment.

Joined
May 19, 2022
Messages
70
I need some advice on how you guys think I should get a thermocouple into my forge for HT. I know that the thermocouple should be protected from the direct flame for best readings and in some form of chamber. But I do not know the best way to get it in there. Currently I have mine just attached on the inside of a small square tubing chamber, but the wires are in the front end of the forge which led to them being burnt up and now it doesn’t work to well (I expected this lol). So what’s a better way to get just the node? Probe? Not sure what to call it but the thicker rod that is meant to do the reading, into the chamber without the wires being in the forge haha. Also I’m in search of a new thermocouple and not sure what to look for, links would be appreciated, I already have the part that reads the temps. As you can probably tell I know very little about these things so please help me learn. Here are pictures of the forge
Before anyone worries about the refractory it’s getting recoated today lol.
 
Drill a 3/4" hole (or one that is the size of the ceramic sheath you have) through the top and put a ceramic sheath in it with about 1" sticking out inside the chamber. Use some satanite or other refractory to seal up the sheath to the shell and inner refractory. Put the TC in the sheath. Shorten the leads as needed so they stick out about 2-3". Plug the sheath up with some kaowool. Connect the TC leads to the PID or reader with type K TC wire.
 
Stacy, would you drill that hole about mid-way between the two burners? Or, other location?
 
Drill a 3/4" hole (or one that is the size of the ceramic sheath you have) through the top and put a ceramic sheath in it with about 1" sticking out inside the chamber. Use some satanite or other refractory to seal up the sheath to the shell and inner refractory. Put the TC in the sheath. Shorten the leads as needed so they stick out about 2-3". Plug the sheath up with some kaowool. Connect the TC leads to the PID or reader with type K TC wire.
Would the TC need to go into the square tubing that the blade goes into? If so does the fact that some of the rod is going to be in the forge and not in the tubing affect the temp readings? Or does the tc only read at the tip? Sorry for so many questions I just wanna get it done with minimal issues ahah.
 
Just a heads up: If you connect the old TC with type K wire as recommended by Stacy, pay attention to get the wires connected correctly. There are two different types, and you need to connect them to the matching one on the TC leads, otherwise you create another junction that is affected by temperature changes. This is not immediately obvious as the TC reader will likely give you somewhat plausible temperature values when you fire up the forge. You can test your wiring using a hot air gun on low or a hair dryer to gently heat up the wire connections you made. When wired incorrectly, this will change the temperature reading and you have to swap the wires. If you use a new TC with long enough wires to reach the reader, you don't have to worry about any of this.
 
My thoughts are, that a longer thermocouple might help your situation. It sounds like you may have gotten it figured out though between Stacy's suggestion, and learning where the thermocouple reads the temp.
 
The best way to determine the wires is a magnet. On type K TC wire, the negative lead is red and magnetic. The positive lead is yellow. Use the magnet to find the corresponding lead on the TC.

You can shorten the TC leads as much as you want. It is only the tip that reads. About 6" should be plenty far from the forge shell in most cases. If your forge is well insulated, you can mount the TC block right on the forge.

Putting the TC in the muffle tube is the most accurate way, but may need some figuring out. The spacers will need to go all the way outside the tube. You actually won't need the sheath if using a muffle pipe/tube.
 
Last edited:
I think welding a cover and drilling a hole in it on one end would be best?

Idk if you've tried that yet, but that seems to be the popular option for doing this.
 
Could you elaborate a bit on your method? Welding a cover where?
You close one end of the muffle pipe off with a plate you weld on. That end will go towards the back of the forge (obviously, so you can put stuff in the front still). Then drill a hole in that big enough to fit the probe through. Could also drill holes and screw the base of the thermocouple to the pipe depending of yours can do that.
 
The only useful placement for a thermocouple in heat treating is to stick the tip under the blade in the forge. This will tell you the temp of the blade, not the forge. Don't bother drilling holes, just sit it in place for temporary use. Eventually you'll learn the color/temp relationship and won't need the thermocouple. Turn off the lights, because proper temp for HT is pretty dark compared to forging.
 
"...You close one end of the muffle pipe off with a plate you weld on. That end will go towards the back of the forge (obviously, so you can put stuff in the front still). Then drill a hole in that big enough to fit the probe through. Could also drill holes and screw the base of the thermocouple to the pipe depending of yours can do that...."

That is called a closed end muffle. It works very well. Most folks just slide in the TC from the open front side. Be sure to bend the TC leads down at 90° where they exit the muffle to get them and the TC wires connected to it below the hot exhaust gasses. A TC with extra-long 8-gauge leads and ceramic bead spacers is a good choice here. Obviously, don't let the TC leads short against the muffle walls or end. Use ceramic bead insulatirs/spacers.

Here's how I used to do it before I switched to doing HT in the oven.
A closed muffle is easy to make from any piece of non-galvanized square tubing. To close it up cut some pieces of scrap soft firebrick so the fit in the ends.
This is basically a closed muffle. This makes for better heat distribution and less decarb. Some folks put a chip of wood in the closed muffle.to burn off the oxygen, but I don't.
A piece of heavy walled 3X3" square tubing makes a perfect muffle. If your forge and blades are smaller, try a 2X2" square tube. The real thing you want is thick walls. The muffle should be a little longer than the forge, with the muffle's inner chamber about the same length as the inner forge chamber. Make sure there is still plenty of room around the muffle for the hot exhaust gasses to exit the forge through the ports (see notes below). This is very important with a venturi burner, as backpressure will cause it not to burn properly.

I made my muffle from very heavy walled 3X3" square stainless tubing (scrap yard find). I tightly fitted a piece of soft firebrick on one end and made one that I could slip in and out of the other end. The TC wires were poked through the fixed end and bent downward as noted above. The TC tip was about 1/3 down the muffle I figured that this is the area the blade edge would be located, and where I wanted the most accurate reading. I put the blade in the muffle and sealed the open end with the firebrick plug. The TC is connected to the PID controller. This gave me the ability to do a forge HT with a 10-minute soak and still have good temp control. I don't use it anymore since I switched to doing almost all HT in the oven.

TIPS:
Raising the muffle about an inch off the forge floor is a good idea. This allows the bottom of the muffle to heat up evenly. A couple pieces of 1" thick firebrick at each port works well.

When building a new forge, planning for a muffle is wise. Get the muffle tube before making the plans and building the forge. Make the port height 1.5" more than the muffle and the port width 1" more. This will allow 1/2" clearance on the three sides with a 1" firebrick riser under the muffle. For a 3X3" muffle that would be a port of 4.5"HX4"W. If this port size is bigger than you want when forging, close it off as needed by stacking some firebricks at the ports. You can also cut firebrick inserts to make the ports smaller.

When using a muffle, pre-heat the forge and muffle for a good 15 minutes to get everything fully soaked before adding the blade. Placing the TC in the muffle allows you to adjust the flame so the interior of the muffle is at the target temp.

If using a PID controlled burner, pre-heat for 5-10 minutes manually first to get the muffle soaked before engaging the PID control function. Try and get the forge running close to the target temp. Give it another 10-15 minutes on PID control to get zeroed in on the exact target temp. The muffle slows the response time of the PID and it will take a while to get tuned to the swing time.
 
The term that will get you in spec is Thermowell. That is what separates the Temp Element from the process. Your application would be best served by protecting your sensors.
 
Your protected element should be as far from your flame as possible. If left in intermittent flame or steady, you need to find an absolute reference and find the offset or adjustment.

Just think that you have a temp range of 0 - 2500 d F in your process. Your TC will send a signal of milllivolts because the dissimilar metals of the TC expanding at a different and predictable rate.
This is converted by your controller and the sent as a 4 - 20 ma and that is your signal span of 16ma. Process Range (Deg F) is divided by the signal span, that is almost always a 4-20ma or 3-15psi signal.
Because you can find for every ma (4ma is where you start and is ever present unless you are not receiving any signal. 4ma is where you start to add to find where you are at in your process at any given time 4ma is when any diode will respond in linear fashion) that you add, you have one increment of 156.25 F found in your process.
This can be used to be no more than a reference thermometer or serve as the feedback in a temp loop for a fully automated system. Either the gas valve is so common and cheap or rare and expensive, I cant find the control valve for LP gas. Forge size, anyway. Manual adjustment should be all anyone needs as far as I can tell.
 
Last edited:
Judcargile - Take a look at my two-stage PID controller plans in the stickys. It eliminates the need for a gas-controlled valve. Basically, it regulates the flame between pre-set high and low temperatures and can be within a couple degrees of target. Simple to bui;ld.

I agree with you that finding a place out of the direct flame is important. That is why I suggested a ceramic sheath (thermowell) between the flames on the top.
When doing HT, placing the TC in the muffle (which acts as a thermowell) is the only way to be accurate.
 
Back
Top