Heat Treatment Oven

[Dr.Al]

This is one of three projects I'm going to be writing up in parallel (the other two are described here and here). They'll each be running at very different paces, with two at the design stage while one is being made, but I thought I'd start posting updates on them all at the same time. I don't have space to work on multiple projects at the same time, but that doesn't stop me thinking and planning!

This is the least established of the three projects: the design is far from complete and I haven't started buying any of the materials yet. It's going to be quite a slow-burn project: I'll do a lot of thinking and designing gradually while I do other projects and I probably won't start work on the actual build for several months. Nevertheless I thought I'd start writing about it now: that way I'm more likely to get useful advice on the design while it can still be updated.

This project is an attempt to make a heat treatment oven. I definitely don't need a heat treatment oven: for the sort of work I typically do, I can get by with a blowtorch and a fair amount of patience. So why am I thinking of making a heat-treatment oven?

1. As usual, I think it'll be an interesting project (and that's generally enough of a reason for me!)
2. Most of the stuff I've heat treated in the past has been relatively small (d-bits, plough plane blades etc); however, as parts get thicker and with a longer cutting edge, it can take a long time (and a lot of gas) to get enough heat into the part to get it up to temperature. The time is not ideal as I'm usually squatting on the floor in an uncomfortable position to do the heat treatment!
3. It should allow better control of temperature during hardening (and tempering?) so hopefully I'll be able to get more consistent results when hardening steel. On one occasion I also found that I hadn't managed to get the part hard enough (either due to not heating it enough during hardening or over-cooking it during tempering: I'm not sure which) and I had to start again. It could potentially also allow hardening of other materials (e.g. stainless steel) if I ever want to use them.
4. I'm (optimistically) hoping that a heat treatment oven could also be used as a small furnace for melting brass and possibly bronze. I've got a lot of brass scrap stored in a drawer ready for a time when I can melt it down and turn it into something else. I'll probably end up just turning it into bar stock rather than anything complicated, but either way it needs a furnace of some sort to get it up to its melting point (930°C). Having something with good temperature control is probably a good thing for brass given the fact that zinc boils at a lower temperature (907°C) than brass melts.
5. Did I mention that I think it'll be an interesting project?

My usual approach to heat treatment is to do make a little tunnel from ceramic wool and use one or more blowtorches to heat the part. You can see an example of this in the following photo:

Judging the blade temperature is a combination of looking at the colour and testing it with a magnet. After testing, I usually stick it back in the flame for a bit to compensate for the temperature loss when testing it: it cools down quite quickly. All of that introduces quite a bit of variability.

Of course, I could just build a better frame (at bench height) and carry on using the blow torch, but if I'm going to do that, I might as well go the whole hog and try to build an electric oven.

I'm expecting the heat treatment oven to be a lot slower than the blowtorch approach. However, that extra time will be spent on the oven heating itself up rather than me squatting on the floor holding a blowtorch, so it should feel less like a chore.

My plan is to make it slightly unusual in that I'd like it to work in two orientations:

1. Horizontally, so the door opens like a cabinet door: this would be used when heat treating as it keeps as much heat as possible in the oven when you open the door and also means you aren't looking down into something that's kicking heat upwards.
2. Vertically, with the door at the top: this would be used when casting with a crucible lowered into the oven (putting the crucible on its side would be rather messy!) This is going to involve a few careful design considerations to make sure a crucible sits in a position where I can get some tongs around it.

I'd like it to be relatively small and it's important that it's not too heavy. The oven won't have a permanent home in the workshop (as I have no space) so I need to be able to move it for storage. Also, I'll definitely need to use it outside if I'm doing brass casting (to give any zinc fumes freedom to escape; I'll also be wearing a respirator if/when I do this).

My initial sketches suggest the internal volume will be approximately 154 mm × 154 mm × 200 mm, although the depth (last measurement) might reduce a bit. When casting, there will be some sort of stand for the crucible to sit on (effectively reducing the depth a bit) so that the top of the crucible is accessible for tongs.

This CAD model image shows my current plan for the brick layout. The insulating fire bricks I'm expecting to use are 230 mm × 114 mm × 76 mm, and this layout gives 154 mm × 154 mm × 228 mm interior dimension. The door (not shown) will probably have a protrusion that goes into the opening a bit to improve the heat seal, hence the fact I'm expecting the internal volume to be slightly smaller.

This CAD model image shows my current plan for the back face layout. The door will probably be laid out in much the same way. In theory, I can cut the three small bricks out of one full brick, but I'll buy extras in case it's necessary.

That's about as far as I've got so far: there's still a lot of designing and planning to go. It'll be a fair while before I get started making anything so I've got plenty of time for planning.

 

[hotponyshoes]

and it's important that it's not too heavy.

I think you are at 15kg with just bricks already..

 

[oldbikedavey]

Electric is better than gas, as the atmosphere composition is more controllable, as is the temperature.
High grade insulation is essential, and all furnaces are to be regarded as fairly high maintenance items, if not actually disposable as such.
But, as ever, there's more than one way to skin a cat.....

 

[Dr.Al]

I think you are at 15kg with just bricks already..

Yes, I think that's about right. I'm hoping to keep it below 30 kg overall (preferably quite a long way below!): I know I can lift that (not for very far, but that's okay).

I'll use some sort of separate-able hinge for the door, so worse case I can take the door off when moving it: that should help a little with weight.

 

[Dr.Al]

Electric is better than gas, as the atmosphere composition is more controllable, as is the temperature.
High grade insulation is essential, and all furnaces are to be regarded as fairly high maintenance items, if not actually disposable as such.
But, as ever, there's more than one way to skin a cat.....

Thanks: all advice welcome!

 

[oldbikedavey]

Thanks: all advice welcome!

The one that I have is a Flamefast LN1000 gas kiln, which is capable of some fairly useful work, but isn't particularly good at heat control without much dedicated attention and mollycoddling.
It will melt most bronzes, aluminium, and zinc=based casting alloys, but really struggles with copper at 1060C.
Pack -hardening, though, it does quite well, if the strongly oxidising atmosphere is planned for and taken full account of.

 

[wyn]

I see the bricks are described as insulating but are more likely refractory.

They do start to break down and fall apart after a couple of heat cycles.

 

[Dr.Al]

I see the bricks are described as insulating but are more likely refractory.

They do start to break down and fall apart after a couple of heat cycles.

Really? Only a couple of cycles? Everything I've read about making furnaces & heat treatment ovens recommends these "23" grade insulating fire bricks. I'm really surprised that they would fail so quickly.

 

[oldbikedavey]

Really? Only a couple of cycles? Everything I've read about making furnaces & heat treatment ovens recommends these "23" grade insulating fire bricks. I'm really surprised that they would fail so quickly.

It's furnace stuff.
Suspend disbelief, or go mad....

 

[jsf55]

Refractory bricks like being warmed up gently and hate being left too go cold without reheating gently, I've seen full reline on a furnace fail in weeks rather than the promised year !

 

[Pond frog]

If you’re talking about the lightweight white Insulating bricks they will definitely last more than a couple of cycles.
They do crack fairly easily but that doesn’t really matter it won’t effect their performance, my only concern would be your roof bricks are only supported at one end which may be a problem if they crack.

 

[Dr.Al]

If you’re talking about the lightweight white Insulating bricks they will definitely last more than a couple of cycles.
They do crack fairly easily but that doesn’t really matter it won’t effect their performance, my only concern would be your roof bricks are only supported at one end which may be a problem if they crack.

Thanks for the advice. I'll have a ponder and see if I can come up with a better layout

 

[Misterg]

If you’re talking about the lightweight white Insulating bricks they will definitely last more than a couple of cycles.
They do crack fairly easily but that doesn’t really matter it won’t effect their performance, my only concern would be your roof bricks are only supported at one end which may be a problem if they crack.

+1

The bricks will be fine, but find a layout where they're self-supporting, rather than dangling in fresh air.

If electric heating, you may need to add a layer of ceramic board / blanket to reduce power usage or achieve the temperature you want. (Make sure any steelwork is on the outside of the insulation.)

 

[daleyd]

Ooh watching this with interest! My day job is industrial controls, a lot of it for thermal processes.

I’d be happy to donate a few bits and bobs on the controls side from things I’ve “acquired” (ahem) over the years, pid controllers, contactors, SSR/thyristors etc if they are of any use to you.

I’m not really an expert on the mechanical/insulation side of things but I’d agree with @Misterg comments above. For a small furnace it’s probably going to be fine, obviously for first run do it on a slow ramp up to dry it out slowly to avoid cracking.

 

[Dr.Al]

Ooh watching this with interest! My day job is industrial controls, a lot of it for thermal processes.

I’d be happy to donate a few bits and bobs on the controls side from things I’ve “acquired” (ahem) over the years, pid controllers, contactors, SSR/thyristors etc if they are of any use to you.

Wow, that's very kind of you, thank you!

I'm still very early in the process of figuring the design out (although there's a written-in-advance post about the control strategy coming later today...)

I’m not really an expert on the mechanical/insulation side of things but I’d agree with @Misterg comments above. For a small furnace it’s probably going to be fine, obviously for first run do it on a slow ramp up to dry it out slowly to avoid cracking.

 

[Dr.Al]

Following the advice above, I'm going to revisit the brick layout to make sure that the top layer of bricks is fully supported rather than relying on a glue joint. In the meantime though, a bit of discussion of the control system might be of interest.

The diagram above shows a simple circuit diagram for the control system.

The oven will be controlled by a PID (proportional, integral, derivative) controller connected to the heating element via SSRs (solid-state relays, usually in the form of two back-to-back thyristors, often with an opto-coupler driving them, which I've omitted in the circuit diagram for clarity). In theory, one SSR would be sufficient; however, the typical failure mode of most SSRs is as a short-circuit and therefore it is safer to have two. The PID controller will read the temperature via a K-type thermocouple (K-type thermocouples have a maximum temperature of 1350°C) and control the current through the heating element in order to adjust the temperature to match a set-point.

Separate power and heating element switches are included so that the power to the PID controller can be enabled, the PID controller set up and then the heating element turned on when ready. An interlock is used to prevent the heating elements (which are connected to mains voltage) from being live when the door is open.

Choosing a PID controller is an important decision. As I see it, I've got four options:

1. Pick a basic PID controller (circa £15–20) and accept that it can't do anything fancy like controlled temperature ramps.
2. Pick a fancy PID controller (circa £90) that can do temperature ramps / soaks etc.
3. Pick a mid-range PID controller (perhaps £30–40?) with an RS485/Modbus input. This would work much like the basic controller, but I could potentially add an external microcontroller to send commands over RS485 and implement temperature ramps / soaks myself if I ever felt the need.
4. Implement the whole lot from scratch using a microcontroller and a home-designed circuit board.

I'm an electronics engineer by trade, so #4 is entirely plausible but it feels far too much like the day job and I doubt I could make that circuit board for less than the price of an off-the-shelf mid-range PID controller.

The third option (RS485 controller) is the most tempting of them all, but all the PID controllers I've found that have an RS485 input are either up at the price of the ramp/soak controllers or they can only handle temperatures up to 400°C.

That, unfortunately, leaves me with a choice of a simple, cheap controller or an expensive complicated one. At the moment I'm leaning towards the basic PID controller: if I find I need temperature ramps / soaks in the future, I can always replace the cheap one with a more expensive version at a later date.

I should note that I wrote all of that before @daleyd made his kind offer; that offer might change things a little!

 

[daleyd]

Following the advice above, I'm going to revisit the brick layout to make sure that the top layer of bricks is fully supported rather than relying on a glue joint. In the meantime though, a bit of discussion of the control system might be of interest.

The diagram above shows a simple circuit diagram for the control system.

The oven will be controlled by a PID (proportional, integral, derivative) controller connected to the heating element via SSRs (solid-state relays, usually in the form of two back-to-back thyristors, often with an opto-coupler driving them, which I've omitted in the circuit diagram for clarity). In theory, one SSR would be sufficient; however, the typical failure mode of most SSRs is as a short-circuit and therefore it is safer to have two. The PID controller will read the temperature via a K-type thermocouple (K-type thermocouples have a maximum temperature of 1350°C) and control the current through the heating element in order to adjust the temperature to match a set-point.

Separate power and heating element switches are included so that the power to the PID controller can be enabled, the PID controller set up and then the heating element turned on when ready. An interlock is used to prevent the heating elements (which are connected to mains voltage) from being live when the door is open.

Choosing a PID controller is an important decision. As I see it, I've got four options:

1. Pick a basic PID controller (circa £15–20) and accept that it can't do anything fancy like controlled temperature ramps.
2. Pick a fancy PID controller (circa £90) that can do temperature ramps / soaks etc.
3. Pick a mid-range PID controller (perhaps £30–40?) with an RS485/Modbus input. This would work much like the basic controller, but I could potentially add an external microcontroller to send commands over RS485 and implement temperature ramps / soaks myself if I ever felt the need.
4. Implement the whole lot from scratch using a microcontroller and a home-designed circuit board.

I'm an electronics engineer by trade, so #4 is entirely plausible but it feels far too much like the day job and I doubt I could make that circuit board for less than the price of an off-the-shelf mid-range PID controller.

The third option (RS485 controller) is the most tempting of them all, but all the PID controllers I've found that have an RS485 input are either up at the price of the ramp/soak controllers or they can only handle temperatures up to 400°C.

That, unfortunately, leaves me with a choice of a simple, cheap controller or an expensive complicated one. At the moment I'm leaning towards the basic PID controller: if I find I need temperature ramps / soaks in the future, I can always replace the cheap one with a more expensive version at a later date.

I should note that I wrote all of that before @daleyd made his kind offer; that offer might change things a little!

I have a selection, I’ll dig them out and you can see which one you prefer. I think they all have the ability to do programming (ramps, dwells etc).

 

[Kram]

Consider the bricks are insulating and when 3kw of power is applied, you will get a large temperature difference from the element to the surrounding furnace. The element can easily run at 1200c and locally melt the refractory.
A surface coating of alumina grit and sodium silicate binder should make the bricks more durable, as would refractory cement coating.

I used a basic REX-C100 PID from amazon for my furnace and works well. I havent yet needed precise multi stage ramping that I couldnt do manually.

I would suggest adding an SCR controller in the design after the SSR's to give adjustable power as its quite difficult to get elements exactly right. This lets you put a slightly more powerful element in and adjust the duty cycle. It can also be used to limit power and give a slow ramp up. If the element fails and a section shorts out, you can limit the power and continue - as in one of my brass casts where I spilt slag on the element. The import SCR units seem to have leakage to case so should be housed electrically isolated from the contol box housing.

Consider how the element will be retained, as I have blown several where it has fallen out of the slot. With @Onoff 's help we designed a spiral 3d printed former for the refractory clay element housing but I have not got around to it yet ( working space issues).

 

[Kram]

Also, I'll definitely need to use it outside if I'm doing brass casting (to give any zinc fumes freedom to escape; I'll also be wearing a respirator if/when I do this).

After varous brass castings, I found it is important with a gas or fuel fired furnace.
With electric the peak temperatures are much lower, so there is almost no fuming. It would not need much local ventilation to allow indoor/garage/shed use.
I have used mine without a filter mask as it was generally unnessasary with no visible fume. Zinc is very visible white smoke.

 

[Misterg]

In theory, one SSR would be sufficient; however, the typical failure mode of most SSRs is as a short-circuit and therefore it is safer to have two.

FWIW, the common arrangement is a second thermocouple in the kiln connected to a separate over temperature trip that knocks the power off. It offers protection against T/C shorts, controller faults, etc. as well as SSR failure.

Worth having a programmable controller, I would have thought.

Personally, I wouldn't put sodium silicate anywhere near it - it will start to dissolve those bricks as a fairly low temperature (<<1000°C) and generally make a sticky mess. I really don't think you have to worry much about those bricks - they're easily abraded and become dusty, but put up with a lot of abuse otherwise. (Technical manager at a glass works for ~12 years.)