Printable View
This has been a real tough one. Here is part one of my home made electric steel heat treat oven.
I have researched loads of different designs and builds. Here is the one I believe is best suited to guys who have limited equipment and avoids the over engineered models out there. Tell me if you agree/dont agree.
https://youtu.be/5bEWFppRebg
-Thank you for a well made video! Nicely presented considerations and executions.
Also, thanks for the crucial warning regarding the dust produced and the necessity for a face mask.
Being in a "Furnace-planning mode" myself, I've just got a few tips:
1) Blocks can be soaked in water before cutting, (temporarily) keeping the free dust down.
2) Kanthal A-1 wire has higher temp, longer life, (higher prize) but it also assumes:
3) Using the "26" type of blocks is absolutely recommended, as their alumina content is higher than lower grades
(which can be seen from the Kanthal Handbook
https://www.kanthal.com/globalassets...41-b-eng.pdf):
Attachment 35306
4) As they also simply say: "Water glass as a binder in cements must be avoided."
I guess they mean avoided entirely in the oven, as the fumes seems to deteriorate the Kanthal wire over time.
Which really is a PITA, as water glass/ sodium silicate is convenient, readily available and cheap.
Just my 2 cents, and YMMV.
Eagerly awaiting your next project videos!
Johan
Hi Johan,
That's brilliant, thanks so much. I didn't even know that what I was mixing was known as liquid glass. As soon as I read it I realised. Getting a bit slow with age.
I wish I had known that before I did it, but I guess I will now be able to observe its effects! At least my design allows for an easy replacement of the coil. Also great tip about soaking the bricks.
Thanks again, Nick
Sharing is the way to go forward - none of those tips were originally my own.
I can merely read and have a decent memory (for my age).
AFAIK - the need for high alumina content of the "26" bricks (as to compared with lower temp grades),
are that the content forms a greyish Al2O3 protective oxide cover on the A-1 wire when glowing,
thus prolonging the life span of the wire.
This is particularly true when running the oven/ kiln up to and above 1300C,
where the wire really begins to suffer, and where the sodium silicate would become more detrimental.
Water glass is usually one of my go-to materials (as a cement binder, water & fire protection, glue etc),
and it would've darned useful for (cheaply) reinforcing these "26" series fire bricks...
When my own melting foundry/ heat treatment facility (4 kW, 400 VAC 2-phase) eventually will be built,
I guess I'll have to live with the bricks' brittleness w/o any Na-silicate mortar,
having a thin stainless plate as a protective bottom?
Without intention of hi-jacking your thread:
I haven't actually found any suggested mortar for Kanthal A-1 & "26"-grade bricks anywhere...
-Any other readers here that have done so - please pitch in!
Cheers
Johan
One good thing is that my oven will never need to go as high as 1300C, Probably 900C max.
Johan, maybe you can help me with something. I know I bought the wrong thermocouple and I am waiting for the correct one to arrive.
But I set my oven to 600C and put an old file in there. When the oven got to 600C, I opened the oven and the file was bright red hot, and my infrared thermometer read 800C.
How can it be that wrong? Any thoughts?
Thanks for your confidence, hope I won't botch it!:)
Short answer: Trust the thermocouple - be wary of the IR thermo until you've figured out and applied this:
// EDIT: You wrote "the wrong thermocouple" -
did it max out at 600 C and the oven trampled on upwards, or what's the couple's upper limit?
I guess tool steel shouldn't be bright red at 600... Better recheck that before anything else of the text below...
Different types of thermocouples has different output voltages:
Attachment 35317
So having a wrong type could fool your controller into thinking it's 600 when really at cherry red 800-ish C...
For instance - plugging a "K" type into a controller set for a "J" type would perhaps be consistant with your result?
(Just follow the 33 mV line from the yellow "J" line (600C) to the green "K" line and read its temp below)
Sorry - I'm just about to hit the sack, so I wasn't really bright and fast enough...//
AFAIK: IR thermometers work by emission, and different materials have different emissivity values, kinda logical when you think of it.
Remember the black body/ polished body relative heat experiment way back in the physics classroom?
Emissivity is only radiated heat/ time unit - no more, no less.
There's one reason (besides sexiness) for black-anodizing cooling fins...
Then - a single, given material ALSO shows different emissivity att different temperatures as well - making it harder to measure correctly.
Also logical - starting at black at room temp, then heating it to fiery red - different rate of heat radiation/ time unit from the same piece.
For instance - I tried (in vain, of course) to use my IR thermo (1300C tops) for reading down into -80 deg C (CO2 extinguisher) a few weeks ago.
I quick-searched. and found some hopefully decent info regarding these matters (I'm not plugging their brand - and physics dont care for shareholders):
https://ennologic.com/ultimate-emissivity-table/
https://ennologic.com/emissivity-inf...eter-readings/
So for a proper readout - you need to know the emissivity of your chosen material, at the ball-park temp.
Personally, when it comes for my own Al alloy melts - the dross kicks my 30 quid fancy-pants IR's readings all over the place -
so I instead got a 5 quid cheapo thermocouple dipper/meter kit from China for consistent readings before the pour...
Hope this helps - guess someone more insightful reader here will correct me - and that's how I learn too!
Cheers
Johan
Nick,
You can do a sanity check with a piece of carbon steel and a magnet. Once the Curie point is reached the steel will become non-magnetic. This should be around 770°C. You can also set the controller to 660° and put some aluminum in a crucible. The alum should be just starting to melt.
A few points worth mentioning. In my experience with building ovens like this, the thin Kanthal will be susceptible to breakage. Once heated the first time it becomes very brittle. I would recommend at minimum 18 ga. 16 better, 14 better yet. Of course resistance (ohms per ft./m) drops significantly when moving to the heavier gauge. I have a Kanthal A1 spreadsheet on my blog page here. You may cover this a the another video but, make sure to double or triple back the wire on the leads exiting the kiln. This reduces the resistance and therefor reduces heat outside as well as makes for a better electrical connection. Also, the elements will sag when hot. Make sure they are stapled into the ledges. I will make a couple dozen from the Kanthal wire in a J shape and press them into the firebrick to secure the element. It would also not hurt to have a door switch that kills the power when you open the door. 220 VAC isn't fun to touch when extracting a hot blade. ;-)
Thanks for sharing!
Dan
I know it will hurt, but this is a rubbish design, and not because of the effort you went to in building it, but because the design demonstrates that you never understood how heat is flowing, through the walls and (closed) door, out of your furnace. You do need a more than nodding understanding of Fick's laws to understand the spread of heat inside furnace walls and thus the resultant temperature distribution inside the furnace. I am not here to belittle your effort, I am writing to provide an explanation for your furnace's behaviour. Once you understand why you get this behaviour, you should be able to think up possible means to improve on it.
First, temperature measurement. Forget about using a pyrometer (i.e. infrared radiation thermometer), because this will produce a meaningful reading only if you measure the radiation coming out of a black hole. With "black hole" I am not referring to the astronomical object, but to a box with a central, large cavity that is accessible only via a small hole. Only this small hole has an emissivity of 1, all other surfaces will have emissivities less then 1, emissivities that depend not only on the material, but also on surface roughness and time. Thus, forget about looking up any emissivity, it does not produce a trustworthy temperature reading without the black hole emitter.
You could also estimate your temperature using Seger cones.
Or you can attach some weight to the tip of a thermocouple and place that at different positions inside your furnace, to measure a temperature distribution chart. If you achieve a temparature variation of less than 50 oC, you have build well. From the pictures you provide on the placing of your heating element, I would expect a spread of 200 oC or more, just what you have observed.
Once you have measured that and understand this is not simple bluster, we can talk again, this time about where to place the control thermocouple, and where to place the heating wires.
Thanks, Dan for pitching in: Nice blog btw.
Thanks for also pointing out the brittleness of wires, and that max temperature degrades with wire dia:
For instance: Kanthal A wire of 3 mm dia (not the average home shop kiln) could take 1350 deg,
but around 1 mm it'll take max 1250.
Sorry, Nick for perhaps being confusing in my previous post, I totally missed the obvious last night...
The discussion on IR thermometer and materials & temp is probably NOT relevant in your case.
I assume by "Wrong thermocouple" you meant you plugged a "K" type into a PID/ controller designed for a "J" type?
Then, IF you choose 600 C as the limit temp, the controller will open the throttle wide open,
until it gets 33 mV back from the thermocouple input.
Fine and dandy IF you'd hook a "J" type up, which will precisely return that voltage at 600 C,
- but a "K" type would give off 33 mV first when it's reached 830-ish degrees.
I guess this was exactly what happened
(but your IR thermometer could have had its emissivity factor off too, but NOT deviate by 200 degrees...)
Hope this helps!
Cheers
Johan
Dan, thanks so much. appreciate everything you said. As it turns out I did do some of this, like stapling and double twisting the end pieces. I did not however install a door switch. When I replace the coil for a thicker one, I'll try and do that. Great ideal.
Johan, thanks so much for all your help. You did not misunderstand. I bought a K type that was only rated to 600. I have taken all your comments on board. Super useful, thank you so much.Quote:
Originally Posted by DIYSwede
Thanks for your feedback. I would be interested to know where I should have placed my element and thermocouple. Although I think Fink's law might be a tad over my head.Quote:
Originally Posted by CanBeDone
Here is part 2 for anyone who is interested. Thanks very much to those who commented already.
https://youtu.be/WAe7YWjv2Vs
I would recommend setting your sights on about 1100°C for a top end temp. Many popular cutlery stainless steels require a soak at or around 1060°C. 13C26, Elmax, 154CM, CPMS35VN etc.Quote:
Originally Posted by anthonyget
Good point. My new thermocouple is good to 1200CQuote:
Originally Posted by DanCom
Thanks, Nick for vid #2 !
Now, I might seem a little OCD when it comes to your "K"-type element:
1) Realizing your ITC-100 has multi-choice possibility of element types:
"Support Multiple Thermocouples and Resistance Sensors (K, S, Wre, T, E, J, B, N, CU50, PT100)"
was the correct type chosen in its menu before startup?
And, if so - what's your take on the 800-ish deg top temp?
Did the element croak at 600 C, or was it still operative, pulsing the output to the SSR correctly?
2) Was the PID, per chance in a "J"-type factory setting - thus achieving the thermal runaway as I suggested previously?
Just curious after all these years - trying to learn, and avoid mistakes!
Small tip: Just adding an extra breaker (2-pole toggle, lo volt and amp) for the SSR pulse control,
to let the coils "cool off" a few secs before opening the door, might be a good idea for wire longevity.
Thermal shock isn't supposed to be something good for A-1, or any other wire type either.
Cheers
Johan
All the low-cost temperature controllers I've had (Inkbird, MyPIN & CXTG) are defaulted to input Type K. The most common type K in this temperature range has a -100° to 1250°C. These are the eBay/Amazon $10 ones. I have learned to keep a spare handy. The high temperatures oxidize the probe's stainless shell over time. Or you can get a ceramic holder which prolongs the life of the probe.
Johan, please don't worry about seeming OCD. I am so grateful for the time you are taking to help me. Genuinely. I think Dan is right, they do default to type K. The correct one should be delivered tomorrow. rated to 1200C. I will start all by testing again but can't thank you all enough. I will keep this thread posted as to the outcome. You are all brilliant and generous. Thank you.Quote:
Originally Posted by DIYSwede
Dan, I feel this is a foolish question, but please bear with me. I am still worried about the PID showing accurate temp. If I place a blade cold in the oven at start up, and set the temp at 815C for 1070, should I expect the steel to have reached that temp at the same time as the oven, or does one have to wait for the steel to heat up? And if so and if the steel reaches the critical temp, how long does one leave it there? In my forge I wait 30 seconds after losing magnetism, but I can't keep opening the oven. Any advice would be greatly appreciated. Thanks in advance. NickQuote:
Originally Posted by DanCom
First: Why use an electric furnace, since an electric furnace is so much slower in heating up than is a gas-fired one, or even a coke-fired forge? To me, there is only one answer, and that is temperature control. If you build a furnace to heat treat knives, you would want an area inside your furnace that deviates no more than 10oC from the set temperature. A PID controller has no problem controlling a temperature to within 1oC, but does so only if the thermal conductivity of your furnace material can be disregarded, i.e. is large enough that the furnace has the same temperature anywhere, inside and outside. Since this is not the case, this means you have no freedom to choose where you place your controlling thermocouple. Its tip has to be where the furnace is hottest, i.e. right next to your heating wire, and somewhat further back than the middle of the furnace, since a furnace is bound to be coldest on the side of the door.Quote:
Originally Posted by anthonyget
It is also a good idea to position the tube, inside which the thermocouple rests, parallel to the longest extent of the furnace, since accurate positioning of the rather flexible thermocouple is not that easy, and why make it difficult to yourself?
Second: Wikipedia is too restrictive when it describes Fick's laws as laws for diffusion only. Actually, the equations are valid for all sorts of transport phenomena, from how many cars can travel on a road, to how a balloon discharges through its nozzle, to - what I am interested in here - how heat travels through matter. A caution, though: it is the simplest possible description of transport phenomena, and as such does not account for things like melting or turbulence.
Because the mathematical description is the same, it is possible to visualize heat transport as a fluid transport:
Attachment 35328
In this drawing - limited by the capabilities of my CAD program - red stands for a fluid being pumped into a vessel (white) that is open at the top. Inflow and outflow pipes are separated by a baffle (also white) so that any fluid momentum is broken up and need not be considered. Under these circumstances, the total volume of fluid in the vessel may be taken as a representation of the heat inputted into the furnace, and the height of the fluid column as its temperature.
That is the global picture, and it assumes that heat (and fluid) travels everywhere at infinite speed. This is, of course, not true. The furnace walls act as insulators, to slow down the flow of heat and that needs to be accounted for.
We do this by representing the furnace not as a single vessel, but many vessels connected in series parallel, and the limited travel speed of heat / fluid by a suitable small diameter of the inflow / outflow pipes:
Attachment 35329
This picture represents a cross-section through a furnace wall, and because the connecting pipes are suitably small, friction inside the pipes degrades the pressure available to uphold the length of the columns (the temperature) for each component vessel.
Third: Under this heading belongs the expansion of what has been said so far into three dimensions. This I will leave unsaid for today, as I will leave unsaid
Fourth: Implications for the placement of heating wires and furnace bricks,
giving you opportunity to mull things over and discover for yourself some of the consequences of what I have written so far. After all, my objective is not to tell you how to do it, but to enable you to find out for yourself.
Somehow - I suspect because my contribution got too long - my pictures got stripped. Here they are again:
First picture: Attachment 35330
Second picture: Attachment 35331
Nick,
I work with stainless and it has a prescribed "soak" temperature and time to hold at to ensure the carbon is in solution. For 1084, I get it to critical temp, wait a couple of minutes and then quench. For O1, I hold at critical temp for 10 minutes, then quench. Each steel has its own recipe if you will. Steel manufacturers like Sandvik and Crucible prescribe soak times based in the thickness of the steel. E.g. hold at x temperature for y minutes per cm of of thickness. This is because the outside of the steel heats faster than the inside. Another step in the heat treatment process is "equalization." This is a rest period to ensure uniform temperature in the piece. As my oven ramps fairly slow, and 'am usually using thin steel (3mm or less), I often skip the equalization step as the steel is roughly the same temperature as the air in the oven. A fast heating oven will have the interior temperature rising faster than the steel. Fast ramping and thick steel will require longer equalization and soak periods.
There's plenty of room for experimentation as each oven will behave in a slightly different way. Placement of the thermocouple will impact the readings somewhat. I place my thermocouple through the bottom as close as possible to the blade rack. Really what I want is the temperature of the steel and not the temperature of the air in the top of the oven. Make sense?
Also, run the Autotune on the controller if you haven't already.
Cheers,
Dan
Thanks so much Dan. So i have worked out that my reading is about 100C out and realised that the tip of my probe was in perfect line with the elements which cant be good. Thanks so much for all the good advice and the best blog. I made your belt grinder by the way. Very grateful for your sharing so much know how.Quote:
Originally Posted by DanCom
Well for any of you who have had the stamina to stay with me, here is the third and final part of my build.
Thank you to all of you who have helped and given me so much help. Not perfect but I usable and will be used a lot.
https://youtu.be/hO76qgb7djs
<!-- BEGIN /var/www/html/homemadetools/protected/modules/zeus/views/tool/postUpdate.php -->
Thanks anthonyget! We've added your Heat Treatment Oven to our Heating and Cooling category,
as well as to your builder page: anthonyget's Homemade Tools. Your receipt:
<div id="blocks"> <div class="block b1 pngfix"> <div class="bimg"> <div> <a href="https://www.homemadetools.net/homemade-heat-treatment-oven-8"> <img src="/uploads/234001/homemade-heat-treatment-oven-8.jpeg"/> </a> </div> </div> <div class="head pngfix"></div> <div class="left pngfix"></div> <div class="right pngfix"></div> <div class="blockover b1 pngfix"> <div class="title"> <a href="https://www.homemadetools.net/homemade-heat-treatment-oven-8">Heat Treatment Oven</a> <span> by <a href="https://www.homemadetools.net/builder/anthonyget">anthonyget</a></span> </div> <div class="tags">tags: <a href='https://www.homemadetools.net/tag/oven'>oven</a>, <a href='https://www.homemadetools.net/tag/heat-treatment'>heat treatment</a> </div> </div> </div> </div>
<!-- END /var/www/html/homemadetools/protected/modules/zeus/views/tool/postUpdate.php -->
I kinda wish that I had seen this thread earlier as I've worked on validating a number of process ovens and probably could have pointed out a few things of note. The first thing to point out is that unless you have active circulation in an oven the temperature in that oven will vary throughout. Even with fan driven circulation it is a real challenge to get the temperature profiles in an oven within spec.
For heat treat ovens of this type active circulation isn't going to happen so you will have to expect varying temperature profiles. You can actually get variations of a surprising amount which leads to the age old problem of where do you put the thermocouple This isn't as easy as you might at first imagine. For example, at first blush you might want to put the probe next to the mass being treated, however until that mass gets up to temperature it will skew your readings as it effectively acts as a heat sink. The bad thing here is that that can lead to over heating elsewhere in the oven. How bad this will be depends upon many factors (one big factor is the mass followed by how close the TC is to that mass) but could lead to damage in high heat areas.
Since heat rises and the items being treated generally lie on or near the bottom of the oven, you may want to consider TC entry from the side walls below the heater element. Ideally your TC is adjustable for depth thus allowing you to find an optimal location. By the way the bricks will also act somewhat like a heat sink, they are of course thermal insulators, but they do take time to heat up thus impacting the oven profile. So you don't want the TC too close to any fire brick.
As for your temperature controller realize that the only thing it knows about is the temperature seen at the exact point of where the TC junction is. Without active circulation it is safe to say that the temperature varies a lot every where else in the chamber. You do get some circulation from convection and air infiltration but air infiltration just makes the temp profile worse. So sealing against air infiltration can do wonders for addressing the high variability in the temperature profile in the oven. In the case of this oven that means sealing the door with some sort of high temp solution.
Hi Johan,
You were so right about a lot of things. I have had to rebuild most of my oven. I wanted to thank you for your advice.
This is what makes these blogs so generous. Thanks again.
https://youtu.be/vKzFF7SRDMc
Thanks Wizard, the learning curve on this one has been pretty steep.
Not wishing to teach my granny to suck eggs, and I hope my comments are relevant. Many years ago, a company called "tor gem" made a solid fuel stove (heater), that had mica windows in the front I used to burn furnacite beads in it, they get very hot, and the result was that over time the mica "delaminated" and went hazy, not important in the stove, I only needed to see that it was glowing nicely, but where you need to see a process, that might be important.
I don't know if this is normal, but if it is you might want to design your door to be able to replace the mica without destroying the build.
I don't have any experience with compact heat treatment ovens, I think the smallest one I ever had any dealings with had about a 3 cubic foot interior space and most of the others had huge induction coils with forced air circulation. However after reading Wizard69's very informative post I couldn't help but wonder if there might not be a way to direct the currents of the internal natural convection at least partially throughout the interior space of the oven by carving channels into the sides top or bottom thereby taking some advantage of the thermal heat sink of the fire brick. Large flat surfaces readily absorb while poorly reflect but smaller angled surfaces or groves possibly even those coming to a peak might have a tendency to redirect the heat at angles creating a slight thermal circulation.
I'm sure there are those who might have tried doing this and it may be well worth the research
Thank you and I appreciate your input. I see all of this as an experiment in many ways so you may well be right. They certainly do get hot. The way i designed it, it is very easy to rebuild the door if all goes wrong. Thanks again, NickQuote:
Originally Posted by old kodger
If that is your approach - have a look at Rick Sparbers furnace, when laid horizontally. To operate it horizontally, Rick would have to fill the groves holding the heating elements with refractory cement, of course, otherwise it would just fall out at temperature, producing a dangerous short. But that would give him also the opportunity to cut the heating element into three, and controlling it with three separate controllers via three separate control thermocouples.Quote:
Originally Posted by anthonyget
If all three controllers are set for the same temperature, the front and rear controller would pump more energy into the furnace than the centre one, producing a more uniform temperature distribution in the furnace . . .
And once the heating element grooves are filled and electrically isolated from the furnace interior, you can safely put a thick stainless steel plate into the central, "uniform" temperature zone to make it even more uniform. And that is where the MEASURING thermocouple belongs, the couple telling you what temperature the furnace has and at which your knives are heat treated. And don't confuse the MEASURING thermocouple with the CONTROL couples! The control couples belong as close as possible to the heating elements, since they control the temperature of the furnace (zones). The measuring couples (plural, if you are as picky about the real world as I am, in contrast to the stories most people try to sell as true) tell you essentially how long you still have to wait until your knife is within the temperature range where you want it to be.
Welcome to the world of continuous improvements !
AFAIK, I've never seen or heard of any other DIY oven maker using this approach of yours. Feel free to put some links up.Quote:
Originally Posted by CanBeDone
Seems like claymakers and glass blowers have survived and been doing pretty well with free, suspended coils
laying in the side wall grooves thru the years? Easy to roll your own, fit, fix, and replace when broken.
Guess the intrepid newbies poking the fiery hot, yellow-white coils with metal rods have already pulled themselves out of our gene pool?
IMHO: To fill the grooves/ Kanthal A-1 coils in a DIY oven with any cement would probably shorten the coils' life, if it would work at all...
Industrial furnaces are a whole different ball park, a whole variety of solid heating rods (and price tags!):
https://www.kanthal.com/globalassets...a041-b-eng.pdf
Then - I wonder what kind of tight temp tolerances you are talking about, and also curious of your very need for those specs?
Personally, I'll be absolutely OK with a no-frills, cheap build, but then I'll only use mine for annealing,
hardening and tempering plain vanilla tool steel and Al, Zn (and perhaps some occasional brass) casting.
The tolerances in my own handling will anyhow be much wider than the temp diffs in my oven.
For resilience I'll add a door "Idiot switch" saving me from even touching an energized coil (also taking the coils' lives a bit further).
I guess Stan is providing some "good enuff" ovens for more discriminating and qualified workshop users than me,
so his HotShot series might even be of interest to other HMT:ers:
https://www.youtube.com/watch?v=LvSRW7NH6Yw
Cheers
Johan
About 50 years ago, when starting to do my master's thesis in physical metallurgy, the local Research Assistant taught me how to build a furnace: take a tube of sintered aluminum oxide, wrap coiled heating element wire around it, and hold it in position by smearing refractory cement around it, filling all the gaps inside and between the coils. For my doctorate at a different university, exactly the same procedure. Ditto, when I went to do research at a third university in a different country. The materials used for thermal insulation were different, however: ceramic wool (Kaowool) on the first two, hollow aluminum oxide spheres on the third.Quote:
Originally Posted by DIYSwede
The most important instruction, however, was on the spacing of the wire: closely spaced near the ends, and much wider near the middle of the alumina tube, to compensate for the extra heat losses that occur at the end of the tube. Even with this, the temperature profile over the length of the tube was pretty pathetic - the zone within which the temperature stayed the same within a degree or two was only a few centimeters long. So, we extended the zone by inserting a piece of metal, utilizing the metal's higher thermal conductivity to extend the uniform zone.
Unfortunately, at the time this was done, there was no internet, and nobody thought of putting up videos on how to construct a furnace. Who would have cared, when the Research Assistant knew how to do it, and where to source the materials? That was part of his job, and he was paid to do just that.Quote:
Originally Posted by DIYSwede
What I learned only much later (when I learned why CORTEN steel can be left to rust without ruining it, and what the Japanese did to use it without paying royalties to its American inventor) is that the refractory cement not only held the heating coil in place, it also protected it partially from oxygen, and from mechanical shock, extending its life indefinitely. Only if we allowed the furnace to heat beyond its design temperature (1100oC) would it fail, even with daily use.
I rather have a furnace that does not need fixing, and to heat treat carbon steel I do not need a furnace that goes hotter than about 1000oC, or, for stainless steel / HSS, above 1100oC . But if you want to melt glass or sinter ceramics, you do have to routinely work above 1200oC in an atmosphere containing oxygen and nitrogen. And that means that your heating elements will not last, unless you go to fancy stuff like silicon soaked silicon carbide rods, which, if I remember correctly, are good for up to 1400oC .Quote:
Originally Posted by DIYSwede
This depends on the kind of steel anthonyget is going to use for his knifemaking. If he follows the most prominent choice on the internet, that would be SAE 1080. Assuming his SAE 1080 is continuously cast and thus contains about 0.05% aluminum, his choice of quenching temperatures would be from 850oC to 1100 oC or even 1150oC. If he were Japanese and would want to use tamahagane (an aluminum-free, unalloyed high carbon steel), his upper temperature tolerance would drop to 950oC. Both of these steels need to be tempered at 200 oC. The temperature tolerance applicable here is only a few degrees, if you want to make everything optimal. Going for optimal tolerance is applicable only if anthonyget were to manufacture knives on an industrial scale, as it is about maximizing toughness (=resistance to breakage at the same thickness). If quantities are small, the safest approach to avoid breakage is to slightly increase the blade's thickness, since the force needed to break the blade is proportional to the fourth power of its thickness. Since toughness is not only dependent on tempering temperature, other contributors to toughness like oxygen or sulphur content will overshadow the effects of temperature on toughness, and annealing temperature tolerance is firstly chosen to maintain hardness, and that means 200-250 oC or even 200-300oC - assuming that the quenching temperature, and time at that temperature, has been low enough not to lead to an increase in grain size, and high enough to dissolve all iron carbide into austenite.Quote:
Originally Posted by DIYSwede
If I were in his place, I would use SAE 1060 (most ready source for DIYers: leaf springs from the scrap yard), assume that it does contain the required aluminum, quench from somewhere between 900-1000oC and skip tempering at all, as the supposed benefits of tempering on toughness are seriously constrained by the steel's inclusion content (mainly in the form of silicon oxides / manganese sulfide mixed phases, and in aluminum bearing steel, calcium aluminum oxides, too. Occasionally, you'll also find copper (from an electric motor that was sneaked into the scrap from which the steel was melted) or tin (from too many beverage cans in the scrap) limiting the toughness of the steel.
If anthonyget were to use a steel with a higher alloy content( =Stainless or HSS), he would have to choose a higher value for the minimum quenching temperature, since his objective is to dissolve not only iron carbide, but also chromium, vanadium and tungsten carbide, if those elements had been added to the steel. But his upper temperature limit would remain the same at 1100-1150 oC, since he must keep the aluminum nitride precipitated, or have massive grain growth with its concomitant drop in toughness. He will most likely notice this toughness drop by fracturing his knife when quenching. If not then, then soon thereafter, as this toughness loss is not averted by tempering.
On alloy steels, tempering cannot be avoided, since the objective here is to re-precipitate the chromium, vanadium or tungsten carbides, as they are the source of the improved properties of alloy steel. And that means not only the tempering temperature, but also the rate of heating must be controlled so that precipitates may be obtained with the right size, the right quantity, embedded in a matrix that does not change over time. And the latter means that tempering needs to be repeated several times until the matrix has a stable composition.
You don't believe in stabilising? This is what is behind JKeetonKnives "Straightening Plates For The Knifemaker".
And lastly, these temperature tolerances are applicable over the whole length of the knife, excluding, maybe, the tang. And we will have to assume that the blade is 200mm long.
What CanBeDone stated in his last post reminded me of something a blacksmith once told me back when I was a teenager. He was noted for making some of the toughest most wear resistant agriculture and oilfield tools anyone had ever seen or used. Of course he used a coal forge and quenching tubs but he did other things as well a lot depended on what he was needing to accomplish with a particular type of steel to add or remove certain properties he would heat his metals to near liquid state pour on fluxes or powdered metal oxides or salts and do what called beat the sponge out of it much the same way ancient ore workers would make pig iron out of ore by heating and beating the ore until it became what was called sponge then continue repeating the processes until they had made steel out of iron.
His tempering methods were all over the scale both in how he would go about it to the media used to control the cool down. Sometimes what ever it was he needed to temper and retain maximum hardness he would weld up a thick steel box or use a thick walled pipe then weld caps over the ends of the pipe with the object inside then spend sometimes an hour to slowly het the container to the color he wanted then pack that in lime or even heated salt for it to cool slowly over night . cut open the vessel to retrieve the item he was making.
Thanks, CanBeDone for your elaborate and clarifying explanations - much appreciated!
I could hopefully follow and understand your reasonings & experiences on furnace construction
and advanced heat treatment of proper, top-notch alloys for a lab or production facility.
I also realize that your suggested embedded coils would have benefits regarding
working lifespan, thermal shock when opening, harmful vapors and temp gradients within.
Without intention of being polemic:
I perceive most DIYers (incl yours truly) neither have the means nor the resources for achieving these optimal results,
we just have to cope with our limited financial, material and cognitive resources to get "good enuff" results:
Like "Mystery metals" from scrap yards, cobbled-together furnaces with unclear repeatability for their results.
I personally embrace each step to minimize all uncertainties and tolerances within the limitations above,
in which I'm obsessed with putting in an unresonable amount of work and search for info & methods.
But I'm still stuck with what I've already got: 2 dozens of Grade 26 insulating firebricks, a few cartons of Kaowool cutoffs,
a cheapo Chinese PID controller, a roll of 1,2 mm dia Kanthal A-1 wire and some steel sheets and angle leftovers.
I just have to wing these pieces together as fast, cheaply and operational I can.
AFAIK; I haven't seen ANY available (for privates) cements suitable for Kanthal A-1 wire,
so my hypothesis for now is to assemble the bricks entirely without mortar/ cement
with the coils suspended in angled slots in the sides, devised so they won't drop.
The lifespan/ results of my furnace will only be exactly as long (or short) as I deserve thru my own choices and use.
The web and YT is filled with DIY furnaces, some better than others (from my layman perspective),
but I've yet to come by a DIYer or even a supplier offering a DIY solution having the properties you suggest.
Until then I'll just continue to "wing my own with what I've got" -
keeping an open mind (and ditto construction) to modifications when needed,
while always remembering (or being informed of that) I'm neither omniscient nor omnipotent.
Cheers
Johan
