PID controlled forge/heat treat furnace setup

[ymber]

I'm getting around to building a new forge and I've got a notion to build one that can hold precise target temperatures so it can double as a heat treat furnace. I want to use a PID controller with an S type thermocouple on the input and a modulating valve on the gas line connected to the controller output. I'm not very familiar with PID controllers but looking at most of the existing solutions for this, people are using solenoid valves connected to the controller through a solid state relay. I don't like that solution because it doesn't allow for throttling the gas, only turning it off completely so oscillation around the target temperature is quite high and you have to deal with the non ignition hazard. I see Denis Tyrell's solution to that and the non ignition problem where he has the low pressure gas bypass around the solenoid valve and his results are decent but I'm sure a modulating valve is a cleaner solution and with a well tuned controller it could reduce oscillation to near zero. I'm currently trying to work out a controller and valve combination to do it with. Most PID controllers I see are relay output but I'll probably need something with DC output to control a modulating valve. It's the valves I'm having trouble looking for. I need something that's 1/2" BSP, can take propane, and takes some kind of control input that I can get out of a PID controller. Is anyone familiar with these kind of setups?

My plan is to put the valve on the main gas line before it splits to a pair of 1/2" T burners. NA burners seemed like the way to go here since the fuel/air mixture is somewhat self regulating as the fuel line pressure reduces which simplifies the whole setup as I don't have to throttle separate fuel and air lines proportionally. I've had good success with my previous 1/2" T burner and I reckon the tuning range on them is good enough to work for what I'm trying to do here.

[Latticino]

I found a bunch of options just googling "modulating propane control valve".  However, this is a pretty sophisticated setup for someone who has never played with a PID controller.  Make sure you get one with autotune and that your gas valve fails in the closed position (spring return).  I worked something like this with a solenoid gas valve on my main gas supply and a pilot burner to keep things running when the main burner cycled.  Worked pretty well for me, but thermal masses of the furnace I was controlling were much higher, which can help stability.

Good luck and be safe.

[ymber]

It looks like what I'm looking for is called a proportional solenoid valve. They're the right size, spring return NC, they have a signal input for partial opening, and they're designed to take fuel gases. There are other kinds of modulating valves but all the ones I can see with propane safe seals are big expensive ones for way bigger fuel lines than I'm working with so it looks like proportional solenoid valves are the way to go. Now I'm trying to work out how the control input works. It says the control signal is PWM but the graph on that data sheet shows a DC analogue signal as the control variable, not a PWM duty cycle. If I can control a proportional solenoid valve from the DC analogue output of a 3 state PID controller that's ideal. Otherwise I might look at running the valve off a microcontroller board and just write my own control algorithm for it. I'd have to connect the thermocouple through an amp circuit or something that way but I could probably make it work. Looking at this stuff I increasingly understand why people just rig a gas bypass.

[Buzzkill]

To me it seems like you're trying to force an analog control on a digital tool.  I've built an electric furnace/heat treat oven with a PID controller.   For most of these setups the relay is either fully on or off, and the time spent on or off determines the temperature rather than trying to use something that holds valves partially open or uses variable current.  I'm not suggesting it's impossible, but it does seem like you're making it far more complicated than it needs to be in my opinion.

If you have your Frosty T burner(s) properly tuned, then you never have to adjust the air.  All you have to do is change the gas pressure and it will pull in the right amount of air.  A bypass setup with the lowest functional pressure for the burner and then a solenoid controlled main line set for somewhere near the max pressure you plan to use is all you need. A PID controller with an autotune feature will figure out the rest of it.   Of course if you change the configuration of your forge or use significantly larger or smaller openings at the front or rear you may have to recalibrate.

If you want to ensure that you aren't dumping fuel into your forge when the flame isn't lit you can use flame sensors much like commercial furnaces do.  I'm not sure, but a PID controller may even have a setting that can detect that no increase in temperature is occurring and shut off the fuel.  I only learned the things I needed to make my oven function, so I'm not familiar with all the possibilities.

[Latticino]

Basically what I used on my temperature controlled furnace, except I used an independent pilot at low output and a solenoid main burner.  Works well, but you will burn your type K thermocouples out pretty quickly if your forge stays up at welding temperatures for long.  I used to use a high alumina thermowell to protect mine, though they don't always like to thermocycle.  

[ymber]

I think I'm settling on the low pressure bypass design with a open/shut valve on the main line. Looking at some of the builds people have done like that they're getting +-5C at welding temperatures from it and I reckon there's more precision to be had from it with good tuning. Thanks guys, changing the design early will save me a while messing with hardware that isn't meant to do this. Mangling an analogue valve into a digital system described the original plan pretty well.

Now I'm looking at building the forge before I start on the electronics. I want to do mostly bladesmithing in this forge and some smaller general blacksmithing. The largest things it will need to fit are axe heads or large knives and hopefully damascus billets down the line. I'm planning to run a dual burner setup with a pair of 3/4" T burners so I'm thinking 500in^3 is a good volume for the forge interior. That should be able to get up to welding temperature running 15PSI or less of gas from the tank so fuel efficiency should be alright. It's a quite similar design as in the video at https://www.youtube.com/watch?v=QuoKVMZgLgo. I'm planning to use a 12" diameter 15" long water pressure vessel for the shell. With the ends sawn off, 2" thick ceramic wool insulation, and a 1/4" layer of castable refractory over the top of that I get 7.5" internal diameter and 13.5" long for 596in^3 total internal volume. I expect to lose 100in^3 of volume on the bottom for fire bricks to make it D shaped which gives me about 500in^3 in the end. I'm planning to use the MonkeyForge kiln wash recipe since I can't get Plistix in this country. I'm thinking since I want to use this for precise heat treatment as well as forging the dual burner design is the way to go since it will heat the forge more evenly. If I have a blade in a pipe in the forge while it's holding a steady temperature I should be able to get precise soaks for heat treating stainless steels.

That's my plan, does anything there sound wrong?

 

[Latticino]

To allow the turndown ratio you will want for heat treatment (it is really nice to have the capability to turn down to around 1250 deg. F for a subcritical stress relief pass) I would recommend considering making one of your burners rather small and the other large.  The small burner gets tuned for the lowest temperature you want to run the forge at, and the larger one cycles with setpoint.  This way the smaller one acts as a safety pilot (if aimed correctly) and you only need one set of controls.

I like a thicker castable refractory layer than 1/4", but only if you are using the correct material.  A lightweight, castable, high alumina refractory like Kastolite 30 works extremely well, but I like to see it at least 1/2" thick (if not 3/4") for real structural stability.  Of course you do have a bit more time to heat up, but there is a thermal mass benefit as well.

[Buzzkill]

If your build is based on the video you linked to you should end up with a well-built forge.  The only issue I had is the consistency of the castable refractory.  To me it seemed there was too much water, and that could make for a weaker lining. Whenever I've used Kastolite 30, it is not able to be poured like that when mixed the way I understand is correct. However, it's a small item and we aren't applying a lot of structural force to the lining, so it will probably be ok.  I've never used any type of glue to hold the blanket to the shell either, so I have no idea if there's any real benefit.  Other than that, it appears to be a well thought out, designed, and constructed forge.  I like the removable floor and the sliding stock support - although that ideally would have a handle of some kind to help slide it in and out.  Those tend to get hotter than is comfortable to grab with a bare hand while the forge is in use.

When you say "heat treating stainless steels" what exactly do you mean?  If you are referring to ramping up and reaching prescribed quench temperatures it could/should work well for that.  If you want to be able to temper as well you will probably find it difficult or impossible to maintain temperatures in the 200 to 400 degrees C range with the same forge/burner combination you used for achieving quench temperature.  You may be able to run a significantly smaller burner and use the same forge body, but for me electric is the way to go for tempering.

Other than that I'll just mention that most of us started out with a forge larger than needed because of the mindset of "better to have it and not need it than to need it and not have it."  However, for many of us, that forge sits gathering dust or has been tossed out.  Now when I think about rebuilding/modifying my forge it's always along the line of trying to figure out the shape with the lowest volume that will still allow me to do what I want to do.  If you are going to be making stainless blades 15 inches long you may need that extra length, but it does come at the cost of significant extra fuel consumption.   What you are suggesting isn't crazy big.  It's just a bit bigger than would be recommended for forging.  When you throw in the heat treating aspect you have to decide whether it's worth the cost to have the extra length or to stand there and pump the blade back and forth through the forge to get an even heat before quenching.

[ymber]

For heat treatment I really just want to be able to hold accurate hardening temperatures for a given length of time. I hadn't thought of the tempering issue. We've got an electric kiln in the works already down at the shop, it's only meant for melting aluminium so it won't go up to steel hardening temperatures but it should be good for tempering. I've been looking around information on using a baffle in a forge to get an even heat in a reducing atmosphere for hardening and I reckon it's pretty promising.

I'm planning to build a smaller 300in^3 single burner design first now. When I need the bigger one down the line I'll have plenty of materials left to build it. Going to get building next week. If I increase the castable thickness to 1/2" I get a 6"x12" cylinder, minus floor gives marginally under 300in^3. For one this size would it be worth doing 3/4" thick castable?

[Buzzkill]

The appropriate lining thickness depends on a few things.  If it's a small forge where you are dealing with odd-shaped pieces, and/or you are a little clumsy, then you need a lining that can take a little abuse - which translates into a little extra thickness.  Generally speaking, it's a good idea to have the floor a little thicker than the walls due to the constant abrasion of stock being moved in and out.

The thicker your lining is, the longer it will take to bring the forge chamber up to temperature - which translates into extra fuel cost.  The upside is that if you do long forge sessions with a thicker lining it will act a little bit like a "heat battery" in the sense that once hot it will retain and radiate heat back into the forge more than a thin lining.

So, ultimately how you plan to use your forge determines what is best for you.  If you are dealing with mostly straight pieces where you are unlikely to gouge the walls, then thinner is probably better. This is especially true if you will usually be doing relatively short duration forging/heating sessions. Personally I do not see any real advantage to going 3/4" thick. For the floor that's not outrageous, but even there it's a little on the high side in my opinion.

[Goods]

I agree with Buzz above, but I will add that if you’re planning on using a temp controller with this forge the added thermal mass may help stabilize the temperature making it easier to keep a tight temperature range. 

Keep it fun,

David

[Mikey98118]

It must be added that what castable refractory you choose will effect the values mentioned above. Castable insulating refractory will work far better, with less fuel, than castable refractory used in glass furnaces, for instance. Also, heat reflective finish coatings are just as important as type of refractory used.

[ymber]

I'm planning to use the INS1300-H refractory on the sheet here. The floor of the forge is going to be replaceable fire brick and I'm not expecting the castable to get gouged much. I'm considering using this zircon paint instead of making my own. It's hard firing, good up to 1750C, and resistant to fluxes so it seems ideal. I'll paint the whole inside of the castable and the top of the floor bricks with 1-2mm of it. As far as I can tell it's very similar to the Plistix 900 that people usually use in America.

[Frosty]

Both look good from here the refractory is high alumina so should be resistant to hot flux erosion and the zirconium refractory looks like a good kiln wash. I would definitely use the zirconium silicate wash rather than a home brew!

You should NOT need fire brick for a floor, you can lay a coat of both these products over the current one if necessary. It says so in their literature.

I clean flux puddles out of my forge if they become an issue by putting on a couple layers of leather and wire brushing the forge floor at high orange heat. Dropping the gooey brush in the slack tub removes most of the borax, dissolved brick and scale goop on the bristles.

Frosty The Lucky.