Heat treat oven build advice.

Hey guys,

I'm working on design of a heat treat oven in my spare time and have nearly all the parts. I'm an electrical engineer and will be building all the controls myself. I'd appreciate some advice on which features you like about the ovens you own.

some specs:
- dimms 16"deep x 4.5" x 4.5" (small but big enough for a couple knives)
- Controls are 120/240 so if i find that the oven is too small I can increase the size and power (dont have 240 in my garage yet).
- Approximate wattage is 1KW for 0.187ft^3 ( based on 5kw/ft^3 in most commercial ovens)
- IFB is rated to 2800F
- Elements are kanthal a2
- Controller is a microprocessor based PID with a type k thermocouple as feedback, programming of heating cycles will be done with usb to a laptop in the interim and data will be logged to an sd card (might store the heating cycle programs on the sd for easier transfer to laptop)

Eventually i'll have a list of specs and some schematics, bill of materials etc. I think my total cost will be lower than buying an oven ( salvaged most parts) , but that could be a pipe dream.

Any tips are appreciated.

 

Hi, I built mine in the same general specs in mind. I wound some 18AWG A1; three 19 ohm elements in parallel for about 2100 W. I posted some of the WIP photos on my blog: http://dcknives.blogspot.ca/p/electric-forge.html and http://dcknives.blogspot.ca/p/winding-kanthal-a1-element-for-an.html It's been working good for me.

Let me know if you have any questions.

Dan

 

In making the assumption that you will be heat treating stainless steels, based on my build, my top suggestions would be:

1. Up the watts to around 8kW to 10kW per cu.ft. - these things can take a long time to get up to temperature. The transition from Preheat temp (say 760°C) to Austenize temp (say 1060°C) will be quicker if you have the power. An under powered (or leaky) oven will make this last 300 degrees slow. My temp vs. time plots look more like an exponential curve, that is it kicks ass from 0 to 500, but progressively slows on it's climb towards 1100.
   
2. Buy heavy Kanthal, I'd recommend 18 AWG minumum. This will last longer as heavier wire is less prone to breaking.
3. Make it easy to replace the elements. You will have to replace them one day. It is a matter of when not if. (I am on my 2nd set)
4. Make your staples out of the same Kanthal wire you use for the elements. This will reduce any galvanic issues that can occur when using dissimilar metals.
5. Cut trenches for your elements in the IFB that physically support the elements. Gravity will help keep your elements in the trench. Otherwise, you need a pile of staples.
6. Avoid putting elements on the ceiling. The same idea as #5 above.
   
7. Reduce heat issues: Put a fan in your control box if it's mounted on the oven proper. Also, spec SEW/SF2 wire going to the element leads. These areas can get warm. As you know, garden variety TEW has 90°C rated insulation. Another way to avoid the heat is to build the control box away from the oven. Stand it off an inch or two and you'll be fine.
8. Mount the SSR on aluminum. Be generous with the thermal compound.
9. After the first run, use a propane torch to heat the element red hot before moving, stretching, pulling or bending it. Kanthal ships annealed and is easy to work in that state. Once it's fired it becomes hard and can snap easily if you wank on it without heating it first.

We'd love to see what good mess you get up to.
Please let us know how it's going.

Dan

P.S.
I've got an Arduino Uno with LCD shield and type K thermocouple shield coming in the mail. Planning on making a ramp controller with it. I'll keep you posted. ;-)

 

I definitely second what Dan and Myth have said. The one I have built is 20"x 6" x 6" and it draws 18 amps on a 110 volt circuit. It takes just under an hour to come up to 2200 degrees. I built it from bricks that i cut in an overlapping way so each joint was off set half way through, makes fro a tighter seal. I also wrapped mine in the 1" thick pottery insulation as well. and then in-cased it in 1/8" sheet metal. I would say pay extremely close attention to how the door seals, that's where you would lose most of the heat.

Dan, would love to see the build of your ardruino uno.

 

Thanks for the info guys. I'm putting the finishing touches on the kiln build. Here are some updated details:
- The interior dimensions were increased to 21" x 4.5" x 4.5" to accommodate larger work pieces
- The single element is 19.5ohms and runs at 240VAC, producing around 3KW of power. The power density is around 12KW/ft^3. The element is 18AWG Kanthal A1.
- The SSR I selected is rated for 40A @ 240 continous load, and is mounted on an alluminum heat sink.
- The element hookup wire is SEW2 and is rated for 200C.
- I had originally designed the brick to be interlocking with a removable lid, but this left gaps around the door and I've since sealed it up with mortar. Hind sight being 20/20 I should have just mortared the bricks in place without messing around with the interlocking (as my mortar is rated for 2800F and if you look at some of the big manufacturers they just cement everything together flush).
- Door will be mounted flush to the body (or as flush as I can make it). The door brick protrudes into the oven by a 1/2".
- The arduino uno controller and LCD are going to have minimal functionality but will be expanded upon as I see fit (i've hard coded the ramping program right now just for proof of concept, but there will be provision for programming the heat treat curves directly from the panel).
- The thermocouple controller (sainsmart MAX31855) I bought for use with the arduino is really sensitive to electrical noise. Don't bother getting the sainsmart (based on the MAX31855) one thats relatively cheap off of amazon or ebay as it doesn't have filter caps on the power supply or across the thermocouple inlet (as recommended by the mfg). A better product is the adafruit thermocouple controller (based on the max31855, but has all the caps).
- I'm a sousaphone welder.

 

Good job man. You'll have lots of watts per cube to get where you need to go in a hurry. Please post some pics if you can.

I have a max6675 (Sainsmart TC shield) that I was going to use for a tempering oven controller one day.

Whaa?

Dan

 

ha. must be some filters for language. meant to say im a "horrible" welder

 

Too funny! I was wondering if you were that bad that you could only weld in S shapes.

 

So i've got everything all wired up and have been playing around with the pid loop tuning for a couple hours (super slow response time means cycling isn't very quick). Tonight I was happy enough with the results to warrant a trial up to 1990F. About 10 minutes in and the kiln is at 1000F and I'm happy as a clam, then the temperature started to drop unexpectedly. at around 1100F my $5 "Kanthal A1" element broke. Looks like i'll be winding my own element using the 18AWG kanthal A1 from PSH afterall. I'll be looking at Dan's blog for that tidbit of info.

Also have another thermocouple enroute from PSH that is more suitable than the 24" 316SS probe I salvaged.

 

Oh man that sucks about your element. I have never found a satisfactory way to repair them after they are fired once. They become oxidized, effectively insulated and brittle. I replaced my original store-bought set after about 40 hours of service.

You are running a single 19 Ω element @240 V?

240 V /19 Ω = 12.6 A

Would you consider two 38 Ω elements in parallel? Each leg would be getting 6.3 A. Less likely to act like a fuse if there was to be a kink or pinch in the wire.

Let me know how it turns out.

Dan

 

yup one 19ohm @240vac. i think it was likely a kink/scratch in the element (ie a high impedance spot). mechanically i dont know if i could fit two long elements. I'm going to wind the elements with a little space so i dont have to stretch them.

 

replaced the elements with some 18awg that i wound myself. I ended up reducing the power density to around 10kw/ft^3 from 13. still took around 45 mins to get to 1600f. it was lain so i didnt let it get any hotter. Any tips on how to improve efficiency?

 

Suggestion: Look for places that may be losing heat. Around the door for example. True the mating surfaces as best you can. I used some 80 grit sandpaper and a long, flat block of wood to match the opening bricks to the door.

I had a neodymium magnet latch at first. I noticed a gap would appear in the door around 800°C, so I made a "screw it shut" door latch. I pondered some high temperature gasket for a while, but this new latch is working good.

After much searching, I was unable to find any data on what kind of temperature curve to expect. Mine takes about an hour to get up to 1000°C mark, so you are not too far off.

Regards,

Dan

 

I freaking dropped it

While hunting for leaks I found that the high temperature cement (rated to 2700F) did not adhere very well between some bricks. This meant i could see light from inside the kiln at high temps and couldn't quite get hot enough (1840ish after 1.5hrs). In an effort to fix these issues I removed the electronics box, took the cladding off and fired the oven again. Turns out my roof wasn't adhered very well at all, pretty well the whole perimeter of the lid wasn't making good contact at the high temps and was leaking heat like a sieve. So I removed the elements, lid and side walls and had them sitting upright on my work bench. I turned around for a minute and heard a bang. Both side walls fell off my bench....

But on an up note I have enough brick to resurface the roof, and I can salvage 4 of the 5 bricks used for each sidewall. Time to rebuild and refinish, all the electronics are completed and the programming is done, so its time to be anal about the fit and finish of the bricks (which I wasn't the first time around).

So learn from my mistakes if you're planning on building a heat treat oven.
1) Don't try and get fancy with interlocking bricks. Its a lot of hassle, its hard to get everything to cut nicely, and factory made heat treat ovens (paragon, evenheat, skutt) are just cemented together and they work just fine. (look up the paragon factory tour on youtube).
2) The high temperature cement from Rona (http://www.rona.ca/en/high-temperature-cement-946-ml-p66895006) requires that you soak your soft fire brick. If you don't wet your bricks well enough (and I mean soak them) the brick absorbs the water from the cement quickly and you don't get a good bond. I found that this meant voids in the joints once the kiln was fired, and that those joints were super weak and crumbly. I think the key is thin skiffs of mortar on both whetted bricks then rub them together so that the gap between bricks is minimized. That or buy an IFB Cement from a pottery supply store and mix it exactly as directed.
3) The door seal is a tough nut to crack if you don't spend the time to true up the surface before fitting the door (full sheet of 80grit sandpaper glued to a piece of plywood works ok).
4) If you're inclined to learn about heating element selection read some of the Sandvik (makers of Kanthal) books about heater design (Resistance heating Alloys and Systems for Industrial Furnaces, and Resistance heating Alloys for Electric home appliances). From this you can determine how much you're overloading your elements and what you can expect from them in terms of lifespan. From this information we can determine the suggested ampacity for a 1/4" wrap of 18AWG (with a coil spacing of 3mm) in a groove @1000C to be 7.22A (which is 2A lower than what I'm running through there). Which means my elements won't last as long, but my kiln has a higher power density and should get to a higher temp sooner.

I'm sure there is more to learn as I move forward but thats it for now.

 

Good tips man.

Do you mean that you literally dropped the oven?

All of my bricks are butted or joined on the flat faces. I did a modification when replacing my first set of elements. The roof was looking like it was going to fall apart as it is made up of 6 bricks. My remedy, that seems to be holding, is this: I clamped the roof bricks with bar clamps to make sure it's held together. I took some 1/8" steel round rod and cut it to about the width of the roof, minus 1/2" (approx.). I ground a chisel tip on one end of the rod and chucked the rod into my drill and "drilled" the rod right through all three firebricks. Repeat and you have an simple internal steel framework. The cement seems to be okay on the sides and bottom. The removable roof needed to be more solid that cement alone.

I used this "fireplace mortar" - http://www.homedepot.ca/product/high-heat-fireplace-mortar-296ml/958219
It's okay, but will flake off in open areas, but seems to stay stuck in the joints.


Dan

 

yup it fell and broke. pissed but i can fix it.

 

Bummer.

Keep on brother. It's totally worth it.

Dan

 

I've been meaning to record the temperatures over time while doing a stainless hardening run. I've looked all over the net and nobody wants to publish this kind of stuff.
(Since performance and efficiency can be calculated from this data, maybe the manufacturers don't want to publish theirs.)

The net result is an 85 minute climb to reach the 1070°C (1958°F) mark which is suitable for hardening stainless cutlery steels. I added a couple of notes for fun.



I used a brand-new 1250°C type K probe. Ta (ambient temperature) was cool in the shop, less than 10°C when I did the run.
2180 Watts, 4955 W/Cu.ft.

Hope this helps in your designs.

Regards,

Dan

 

I keep all my data as well in a log , I have found there to be a fair bit of programming and ramp data available under different makers websites.
My forge will ramp to 2300f , welding temp , in about 35 mins, my Paragon to 1950f +\- in 90-100 mins. The Paragons ramp to about 1400F very fast , then designed to creep to max temps