5kW electric furnace build

Hello!

First post here, got recommended the forum over at hobby-machinist.
A little disclaimer on my part, I'm a complete beginner and have no previous experience with brickwork, foundries, melting metals(besides welding), insulation etc etc etc.
Due to this I'm missing several key considerations during the process, learning important things way too late etc.
So be nice.

It all started because I wanted to make a piece from silver.
Absolutely shooting from the hip I bought a small graphite crucible, took a small bucket and wrapped it in insulation(the yellow stuff, for houses, feel free to laugh).
A propane torch for killing weeds was inserted tangentially like some DIY foundries I've seen.

I realized this wouldn't work for long but hoped it would work long enough.
But uh, 10-15 min into it, the crucible is barely dark red whilst the bucket itself is glowing, the insulation is literally melting and even using a second burner directly at the silver, I couldn't achieve a melt.
At this point, with the "furnace" literally falling apart and flames going everywhere, I took my first smart decision all day and stopped.

Took me a few days to get over how badly this failed before I got back on it and started researching.
Even with my limited experience with propane, I already hated it, it's loud and way too much flame going around the place which makes me feel unsafe.
"Ready to install" burners were also quite expensive and I didn't feel comfortable DIY'ing that part.
So my decision at this point in time is to go electric instead.
Since I'm already at the point where I'll be spending some money on this, I figure I'd make it a bit bigger so I can use it in the future for aluminium casting, which I've wanted to do for a long time.
How did I decide on the size? Absolutely just shot from the hip and guessed. Because SMRT.

I've seen enough youtube videos to know that I want those white insulating firebricks to contain everything ... Which are apparently almost impossible to find around here.
The ONE retailer I found, was crazy expensive and with a high minimum order quantity...
Kept looking and eventually found what was basically called "fireproof brick" which seems to be used in fireplaces and decided to just take a chance on trying them along with "fireproof mortar" that's supposed to handle >1000 celsius.

First thing I react to is the weight, these bricks are HEAVY .. But I'm just hoping they insulate well either way, I even bought 50mm/1" thick ones for better insulation(yes, I know now this was a bad call),
The second thing, they're near impossible to shape, I quickly ended up with 2 ruined diamond blades for the angle grinder.
Being stubborn, I managed to find a cheap-ish diamond blade that would fit on my chopsaw, so I both have more control of the cut and also more control of angles etc.
Cobbled together a quick setup.
Bent up some sheet metal underneath to catch the water, used two buckets to contain and catch the water flow, a washerfluid pump, some hose and a computer PSU.
Voilá, watercooled diamondbladed saw.
This actually worked really well, but sure took a while to get all my cuts done.


Either way, eventually I ended up with my bricks cut, testplaced and ready to be mortared together.
Much like a child, I couldn't stop smearing the mortar all over the place because:
A. I had mixed up too much
B. I figured it might help with strength
C. Mentally, I'm a child sometimes and could probably use adult supervision.
Once it was semi-hardened I used coarse sandpaper mounted on a steel plate to flatten each everything out, then let it harden completely.


I did make a lid because I had 2 bricks leftover and figured "might as well".
Reinforced the mortar with some steel rods, figuring it'd help it stay together.
This was a total detour and I won't end up using it.


The heating wire(pre-wound) and everything else arrived after a few weeks, so I connected things together.
Of course put on safety glasses and turned the power on from a safe distance.
While testing I used a SCR-dimmer to control the power input as I didn't want to risk applying too much and burning the wire off or similar.
In second part of the video I went over to 380V with no dimmer and seems to glow nicely with a power input of around 5kW. This is more luck than anything else since despite understanding ohms law I don't quite get how to figure out how much power you're supposed to push into the wire.


Anyway, I ran this up and down between 100 and 200 celsius and I relatively quickly realize that these bricks and insulation have nothing in common.
After maybe 15 minutes the outside is WAY too warm to touch, so I don't even want to know what'd happen at higher temperatures.
Back to the anxious "I f'd up" stage again and start trying to figure out how to solve this.
I could just order white firebricks and start over, but then I've wasted all this work AND I'll be out even more money.
Instead I manage to find ceramic insulation(kaowool-ish) online, which when it hasn't arrived after several weeks turns out to be a scam, at least I got my money back...
So back AGAIN to trying to find a solution.
At this time I got inspired by people trying to DIY firebricks from perlite and cement, so I figured that might be a way forward, as in cover the entire thing with a thick layer of it.
So I ended up making a big bucket from sheet metal, with a thicker aluminium bottom for rigidity.


Trying to figure out how to fill this thing, with the furnace inside, I ended up with a 2 step solution.
3D-printed a mold to make a kind of foot first, filled in with 10 parts perlite / 1 part cement mix.
Pretty interesting mix to work with but no real idea if it's gonna work.
On paper the perlite is supposed to handle 1230c but I know the cement breaks down from high temps so eh.
I think I just pulled and ostrich and sent it.


Since I had quite a generous release angle, it was real easy to get the mold off and then chuck down the furnace.
A second 3D-printed mold gets inserted into the furnace so I could build up some thickness of perlitecement without it falling in/sagging.
Initially I figured I'd have 2 carry handles on the outside but since it's already heavy as ... I instead made a quick under-carriage of sorts and riveted it to the bucket.


Having ordered a huge 100 liter bag of perlite, I figured I'd be set for life, but no.
And mixing everything with a powerdrill and mixer attachment was also pretty horrible work, felt like it took forever.
First pic is how far I got when I ran out of perlite, having to rush to the bigbox store to buy smaller, expensive, boxes of it.
Managed to get it finished up and tried to smooth the surface a bit but not overly happy with it, doesn't need to be pretty I guess.
Once filled up I covered it in plastic wrap to make sure it dried slowly.


Whilst I let that dry for a week I spontaneously started making a new lid.
Much the same process as the bucket itself.
Align and weld strips of sheet metal together.
Roll it in my homemade, garbage-design, sheet roller.
Cut the ring to side and weld it together.
Attach a brim and top.
The idea is to fill this thing with perlite-cement for more insulation.


Obviously far from done, but this is where we're at now, with a wannabe star-wars droid looking thing.
Left to do:
-Fill the lid with crud.
-Hinge/lid lifting system
-Some kind of handle/rod to be able to lift the foot and roll it around.
-Sheet-metal control box for the PID, contactor, fuses etc.
-Implement some kind of safety switch so it turns off the coils if the lid is swung open.
-Watch all the perlite-cement turn into goo as I overheat this thing.


Final words, I really hope this thing will work because I have so much hours into it, but I'm real worried it's going to fail.
Sunk cost fallacy comes to mind.

 

The execution looks pretty good, with the materials being the only unknown in your efforts, you may find the furnace insulation has a lot of moisture to drive off and works fine once dried out. The ability of steam to keep your furnace at lower temps until the water is fully driven off is not to be underestimated.

 

Definitely, just finding the "foundry tutorial" gave me a ton of insight, and regret as to how I've constructed mine.

1. The 8-sided hard firebricks, or super thiccc hot face, sits in the middle.
So basically my layering is from outside to inside 1.2mm steel -> 8-10cm perlitecement -> 5cm hotface with kanthal coils, you can see the grooves in one of the first pictures with the cut bricks.
The grooves are slightly deeper than the coils diameter and angled upwards to contain it securely
Illustration(top view) not to scale, and missing the top layer of perlitecement covering part of the blue bricks.


2. Smallest diameter(flat to flat) is around 190-200mm if I remember correctly.

3. Basically the only purchasing point I could find was Amazon, and the only affordable(very low budget project) option turned out to be a scam.
The Perlite mix is intended only as insulation for the previously shown octagon hard firebrick construction.
I've spent a few more hours searching around afterwards and found a single company tthat MIGHT be able to order it and real refractory for me, but I'm guessing it's going to be expensive if they even want to bother.
Generally companies here don't want to help regular people or small companies..

4. The bag just says "Cement", spec sheet says "CEM II/A-LL 42.5 R" but that tells me basically nothing.
I tried mixing it with some high temperature cement I had but it just wouldn't work, unless I had copious amounts of cement it would just fall apart, and with such a high ratio of cement it had barely any insulating properties.
My hope is that even if the cement starts failing, being packed with so much perlite will keep things in place.

 

If you can get hold of a type of cement called "Ciment Fondu", it is the basis of dense high temperature refactory cement: the insulating properties are not as good as fiber refractory but that can be compensated by making it thicker. Calcium aluminate cement is made in three types: gray, brown and white with brown and white suitable for higher temperatures. White ciment fondu is probably only available from specialist refractory manufacturers. Even the aggregate mixed in with it has to be rated to a suitable temperature too: An Introduction to Refractory Aggregates

 

I've both read your thread(nr2) and seen a bunch of your videos(a while ago, realized this after I signed up on this forum), incredibly inspiring content! Your end results both in fabrication and the casting are amazing and definitely a level I strive to some day achieve.
Real glad to hear that it should work, definitely been a big worry spending so much time on something I'm not confident on working as I have many other things fighting for my time and attention as well.

I doubled the wires starting from just inside the hotface and all the way out, and also twisted them carefully with a drill so I figure the squarearea is probably equal to 3-4 wires as the twist tightened it up a bit.
This will then go into a ceramic connector block on the outside of the furnace, from where I will run copper wire with high-temp insulation(prob silicone, 200c) on it, all the way into the control box.
Mostly so heat-over-time doesn't dry it out and cause a dangerous situation.

Currently with the 380V / 2phase setup I'm not using the SCR at all, I got lucky with the length of the coil so it seems to run at the right power levels(only had it on for like 60 seconds though).
As you mentioned, I might reconnect it temporarily to limit power input when drying it out.
Supergood info on the temperature curve!
Do you set it to 1800F no matter what you're melting or is this specific for aluminium?

I checked my old order and the coils are supposed to be 1.2mm FeCrAl, from what I've gathered this is a bit more fragile but as you say have a higher max temp?

I also did some light reading on DIY crucibles, mainly steel for aluminium, and how it's recommended to get them red hot before use to form a protective layer.
Is this also something you need to do for "DIY ingot molds" or is there less risk there for the aluminium eating up the steel?
The bulk of my aluminium is old wheels so I'm thinking/guessing that it's best to melt these into ingots to clean them up and have more "ease of use" in the future.

 

Sounds good, really appreciate the input!
Making some progress on the lid and hinge.

Suprising amount of work into this, then again it's all from the junkpile so trying to fit things together demands more time.
"Lathed" and milled the right chunk out of a chunk of alu, good candidate for casting but eh, that's why I'm doing this to start with.
Left piece used to be a flat piece of steel and some 40mm round bar.
Milled it to fit inside the square tube and bored/polished it for the right piece to fit inside like a pivot.
Pressfit the "washer" ontop to get more of a bearing surface towards the aluminium piece.
Overkill but also a learning experience.


Going to sit something like this, just as a rough mockup, the square tube is too long currently to have it in the right place.
The idea is to lift the lid, swing it to the side and let it rest on the outer edge of the bucket.
Rotation-wise it will have mechanical stops so it can't go swinging.
Also added a few pieces to the lid, both for "esthetics" and to strengthen the joint between handle and lid.
Welded it completely all around and blended everything down. Partly because it looks good but mostly as practice for car renovations/sheet work.


Sidenote: The casting is feeling super solid so I'm hopeful for it.

 

Little bit more progress, quite a lot of hours for so little visual progress.
Spent probably 2-3 hours on the lathe just making this purely decorative piece for the hinge.
Would've been faster in the mill but wanted to try some techniques I usually don't use.


Previously straight square tube was "pie-cut" and angled out to get the distances correct.
Handle has been cut, angled and reattached.
Further welding and blending on all of this and the lid itself
Handle has also been internally welded inside the lid to make this connection as strong as possible.
Square tube is welded to the "bucket" and I was originally going to reinforce this further but it seems plenty strong for now.


The function is pretty simple, angle the lid up 10-15*, rotate it out to clear the hole and then angle it back down to rest ontop of the bucket.
Overally pretty pleased with how it's come out, really been trying to make things up as I go to improve my fabrication skills.
Normally I would sit for hours in 3D-cad and draw up a final design but I really need to get better at processing this directly in my head.
Next step will most likely be a limit switch to shut everything off when lid angles up and physical endstops so it can't rotate too far.
Also need to fill the lid with perlitecement but I'm a bit worried about chunks breaking off and falling into the crucible. I have high temp mortar at home so I might try using that as a hotface as it's supposed to be able to handle these temps.

 

Cut a hole in the top, with a second metal ring(?), for a plug that can be removed to do quick work like degassing/temp probe?
Weld some spikes to the metal frame to lock the poured cement in place?
Add some stainless steel needles to the cement for internal reinforcement?

 

Sorry I'm late to the show.. Good looking little furnace.. Your fabrication skills really show. Lacking the same, most of what I build is using someone else's discard in a different way than intended. Small aluminum propane furnace.. | The Home Foundry
I used the same layout for the wall bricks, but only used half bricks .. Took a different approach to the top half of the furnace.. I really liked Kellys idea of only handling the crucible once.. Swinging the top half of my furnace to the side allows me to pick up the crucible from the side and pour. Here's some information on perlite composites using perlite and sodium silicate.. No need to use cement to made strong, solid secondary insulation.. PERLITE/SILICATE COMPOSITES FOR HIGH TEMPERATURE INSULATION AND FORMED SHAPES---A SCHUNDLER PRODUCT GUIDE

 

Smoking shoe: Good ideas! Not sure how I feel about having a hole/plug in the top though, it's pretty fast and easy to just slide the lid and since the lid will be rather thick, it'll be hard to reach down unless the hole is almost as big as the opening itself.

Kelly: I just weighed a cupcake of it I made from leftovers. It seems to have landed at around 1 kg per 1.2 liter when almost fully dried out.
The internal volume of the lid is around 7 liters so final weight of the refractory should be ~6kg.

mytwhyt: Thanks! Great link to the perlite stuff! Will have to sit down and read it through properly.

Went at it to sort out both physical endstops and electric limit switch.
Initially I was going to have them on the outside of the rotating part but it looked awful so I spent some time on this solution.
Not sure what the brown-ish round steel disc is, but it's absolutely horrible to drill through.
I suspect it might be stainless because it work hardens real fast...
Either way, the shiny bolthead runs in the larger pocket underneath the hinge, limiting siderotation.
In the larger hole you can just about see a little metal tab, that's from a microswitch mounted underneath, this gets triggered by the smaller pocket on the bottom of the hinge.


Spent way too long on the control box..
Folded up some alu-sheet both for physical and heat protection for the plastic control box.
Behind that is some sheet steel distancing it from the "bucket" itself.
There's also aluminium spacers between the alu-sheet and the control box in order to create an air gap.



Starting to come together though.
Still need to find some insulation cover for the heater wires and a ceramic/high temp connector so I connect my silicone cables to the heater wires.
Once that's figured out I think I can start finally hooking everything up.

 

Ah sorry, I missunderstood.
The bucket has ~45 liters of perlitecement and 10 firebricks.
With the lid, once it's filled in, this should come out to roughly 20kg firebricks and 44kg of perlite around taht.
The steel casing is probably at most 2-3kg.
A final weight of ~65kg seems high, but it was a pain to lift down from the workbench...

As mentioned previously, I wish I had better materials/bigger budget and more knowhow so I could've gotten a lower thermal mass, but that will have to wait for version 2.
The idea right now is to finish this and see how well it works /how long it lasts.
If it breaks or doesn't work well enough I can always just empty the bucket out and remake it with a much thinner hotface and better insulation outside of that, would obviously require a bigger budget/better suppliers though.

 

Real good info, thanks! Slower than I hoped for but as long as it can get up to temp it's not an issue for me.
If melting and casting becomes a bigger part of my time then I will definitely consider rebuilding it for better heat times and most likely do a liftable design like yours, seems super practical.

Gotten a little progress done.
Bent up some stainless wire, hoping it'll reinforce the cement a bit.
Initially I figured I'd just fill the entire lid like this.


But I changed my mind and decided to fill it "in place" so it'd would perfectly conform to the buckets shape.
Put plastic over it and taped in an old headlight to get a dome shape.


Then on with the lid with a new hole in the top and just fill it up.
I tried compressing/filling it as I went but the hole is definitely too small to get enough reach towards the edges.


Resulting in this.
Good packing where I could reach and bad packing everywhere else.
Should probably both have made the hole larger and vibrated everything into place.


Decided to give everything a generous coating of high-temp cement/mortar.
Hopefully this will both seal it up and protect it a tiny bit.
Once this is cured I think everything should be read for some heating tests and start really drying it out.

 

Finally up and running
Large amounts of water/steam kept creeping out as I ramped things up.
Raised it to 100c until it stopped steaming then increased to 200c and it started to steam again.
Wait until it calms down and raise it further, rinse and repeat.
First run I had it up to 600c and let it sit there for a while.
Clearly quite a lot of moisture in the lid as some worked it's way back into the handle which ended up looking like a small chimney.
12h later it was still >65c in the chamber(lid was closed during cooldown) and I started ramping up to 800c with much less steam this time.
After keeping it there for a while it yet again starts dripping from both the bucket and lid, I suspect it takes quite some time and heat to make all the moisture migrate outwards.
With all this extra moisture I haven't bothered measuring heat times, will try once it's all properly dry.


Once it seemed to sit steady at 800c I ramped it up to 950c and figured I'd try melting silver(the initial project I needed to melt stuff for).
While waiting for it to heat up I started making a mold.
Second try at making a mold, first time actually attempting a pour.


Of course ran out of sand and could only fill the cope halfway so had to be extremely careful lifting it off.


Pour itself went pretty good, scary stuff when you're not used to it and accidentally touched the wood with the crucible.
Also got a bit stressed when pouring, so much things going on that I didn't focus enough on pouring at an even pace.


Most of the defects are from the mold even before the pour, had several places of edge damage as I tried to lift out the pattern.


After cleanup, absolutely an adventure and looking forward to doing more stuff but with aluminium!

 

Anything much larger than that casting will need weights on top to prevent hydraulic pressure lifting the cope.