Kelly's Furnace Build Log
Hello all. I joined the forum three years ago. I’ve learned a lot from the folks here. Still have a lot to learn. About a year after I joined I made some major life changes with career and a move of my residence. Shortly after that I fouled up my right hand pretty good so getting my shop put back together after the move was a slow going chore. More recently I’ve been gathering a head of steam and started my furnace build back in December 2015. I’ve been working on it here and there when I can and have it quite a ways along. It’s probably not the typical build and I could probably call it my foundry build as opposed to furnace build as there is a lot more to it than just the furnace itself.
As a little background and explanation to my approach, like many, as a one man shop, I need to be able to move my equipment around, set up, operate, and store it easily. Based upon some of my intended casting projects, it needs to be able to handle up to an A60 crucible…..eventually. I have an A10, A20, & A60. They are all Morgan Super Salamander Clay Graphite. I’m going to work my way up!
Living in the Midwest, occasional small indoor melts in the Winter would be a plus too, and owing to the fact that not everything I try may work, it needs to be easily adaptable and reconfigurable. As a bonus, it would be nice if it could burn out investments and heat treat with it as well.
I still have a ways to go but my shop is coming along nicely. I have many interests and never enough space. Everything needs to earn the space it occupies in my shop so being portable and storing efficiently is an important aspect of the build. There’s a lot of conflicting needs in the preceding ramble but I think I’ve come up with an approach that will accomplish most all of this with a portable unit that occupies a 40” x 24” floor space foot print when stored. This is what I came up and how it sits at the time of this initial build post. Here are some short video clips.
It’s a lift off furnace with a gravity counterbalanced lifting mechanism that relies on and a 2:1 (or 1:2) block and pulley system and a guide carriage on a mast. It also has a linear actuator for the lift that can be used or easily disengaged for manual lifting when being propane fueled or when electricity is not available. It’s a bit more complex than most builds but so far seems effective, easy to use, sets up quickly, and stores well.
The crucible is easily accessible from three of the four sides with either shank or trolley depending upon the size crucible. I’ve also considered a small gantry crane attachment for it. It has several fabricated steel furnace shells that can be attached to the lift. The smaller shell can accommodate up to an A20 crucible (10” bore by 14” tall internal dimensions) and the refractory inserts can be swapped for either electric or propane service.The larger furnace shell can accommodate up to an A60 crucible and at this point have no plans for it to be anything except propane fueled, except for an added electric module section that can be installed for controlled burn out and heat treating. In trying to minimize furnace mass, the radial clearance around the crucible is close but since it’s a lift off, the lift mechanism is very precise, and since the plinths closely positions the crucibles, it’s workable.
The refractory furnace modules are removable and basically freestanding within each section sitting on a steel structure at the bottom of each section. Each successive section sits on the structure not the refractory so the refractory modules basically only need to support their respective weights in each section. I tried to minimize the amount of refractory to keep the mass down so it warms up fast and also to be manageable in the lifted portion without a massive mechanism. It also makes it more practical to wheel around, at least on hard surfaces. Each refractory section can be replaced or swapped in a few minutes with another. This is probably most relevant for the center section of the smaller furnace so it can be swapped from an electric to gas fired furnace. The base and lids have dense refractory hot face so they are used in all configurations but are also easily replaced if needed. I’ll likely add a chimney to the top of the larger furnace to get as much heat as possible channeled above and away from the structure.
The above is a summary. The rest of this (lengthy) post is the detailed build log that I have written piece meal along the way since I started the build three months ago in December of 2015. Looking at the progress pictures and writing the narrative helped me organize my thoughts, improve, and execute the build. My intent was to complete the build and post a reverse build log of sorts. It’s a big project that started from absolute zip, and with the warm weather having arrived my pace may slow a bit so I’d like to post my progress thus far and invite some input. That will especially be the case when I get to melting metal and casting. It’s more of a metal fabrication project to date but there is some IFB work for the electric version of the furnace and I have some interesting things planned for castable refractory module. No CAD models. Just back of napkin sketches and design on the fly for this build.
I built this in my home shop but I run a business which is primarily CNC/mandrel tube bending. We also weld pipe and do some general metal fabrication. We consume many tons of tube and steel per year so I grab the excess drop and scrapped set up parts that are interesting and usable for projects. It’s a fantastic fringe benefit……clean, free, structural metal.
Except the mast and rear guide post, most of the metal in this project is just drop and would have been hauled away for scrap. For the reinforcement hoops on the furnace shell I traded a favor with a buddy to blast those out on his CNC laser table all in one shot though I did need a few more and ended up sawing those out of sheet with a saber saw. The rest is sheet metal I had (mostly) or bought, and then sheared, formed on my brake, and tacked together.
I’m going to post the build log in separate sections. Some are complete, some in progress, and some not yet started. I figured I’d try this approach so I can edit/add to each section over time as it is completed and then just post comments in the later pages as to what sections/post numbers have been updated. That way, maybe it will be an easier read for any folks that may want read it in the future since everything will be in one place.
The major build log sections to follow are:
Other Updates and Refinements
Accessories and Tools
Gizmos, Bells and Whistles
I know how you guys like pictures. I hosted them and they all automatically load. I did resize them all to about 100kb a piece but there are a lot of them so please be patient while they load. Hopefully it won’t plug up you pipes too bad. -Here they come……………………….
Last edited by kcoffield; 07-19-2016 at 06:37 PM.
There was a fellow on Craig’s List offering these locking clamp barrels. He said they were 16 & 55 gallon but the dimensions of thesmaller barrel put them closer to 30 gallon. The smaller barrel was dimensionally pretty well suited for the smaller A20 furnace. The 55 gal was a pretty good fitfor the A60. He only wanted $20 a piece for them and they were like new. The smaller ones had air tight lid seals with locking rings and were so nice Ibought a few of those for dry storage to keep refractory and other humidity sensitive supplies fresh. For the one that gave its life to the 10” bore A20 furnace shell here’s the before and after cut, build up, with strengthening flanges andlift brackets attached, and the finished product before paint. To be sure, it took some effort but the furnace shells are fairly light and rigid. They attach to the sliding carriage and mast clamps with four or six bolts and can be removed or swapped in a few minutes. Here are some pictures of the various subassemblies in the build. BTW, I subscribe to the Henry Ford philosophy when it comes to hardware. Everything (at least everything you would remove or adjust) is 3/8-16 so one wrench will do. And as you will probably gather, I like to fabricate.
Here’s the A60 capable shell built around the 55 gallon drum. It’s constructed the same as the smaller shell above. It has an extra section that can be installed between the base and the upper body which is to be electrically powered (more on this to come). It literally takes one minute to install or remove it. Just set it in place between the base and the furnace body and plug it into the controller.
If I had it to do over again I’m not sure I would use barrels as the starting point for the furnace shell builds. I’d probably just slip roll my own hoops. The barrels are not very uniform in diameter and the larger one varied almost a ¼” in diameter from top to bottom. On several occasions this required an introduction to the business end of my hammer. For the 55 gallon furnace base I actually cut it and patched in a seem that was about 3/8” wide to make it fit the hoop flanges. They’re also quite thin (especially the 30 gallon barrel) walled which makes welding them to thicker metal a bit more challenging. With the time I spent cutting them apart, I probably could have made rolled my own barrel hoops. Given the invested time, it would have been adrop in the bucket.
Furnace Build Update May 15, 2016
The furnace base was oil-canning a bit so I put an end to that.
I wanted changing refractory modules to be quick and easy so I rolled up a piece of 20 gauge mild steel sheet to retain the ceramic wool and also serve as a protective casing.
I don’t have a sheet metal beader or end former (wish I did) but I had a wooden weld fixture left over from the furnace shell build that was pretty much a ready-made buck so I used it to form the ends on the refractory metal skins. I don’t even have a body hammer so I just used a small ball peen hammer and a light touch. They came out ok considering.
Here are a few pictures of how those were formed at the end of post and the finished product. In the future, with the sheet metal skins now in place, the furnace could safely be operated off the lift as a lift-out, pretty much as it sits I suppose, so I wouldn’t necessarily have to swap it onto my lift if I used a smaller crucible that was accessible with tongs.
Last bit of business was installing some gasketing. It’s probably not needed for the routine melts but I was thinking I might do some inert atmosphere melting in the future so now was the time to attend to it. I just used ceramic paper. Some call it refractory cardboard. It’s denser than wool but is still pliable and has some give to it. The place where I got my refractory materials gave me a ¼” thick 3ft by 3ft remnant. Seems like it may work fairly well for electric/aluminum service and has similar refractory properties to the IFB and wool. I was told if aluminum touches it, it will dissolve so it probably would not be of much use in higher temp furnaces but it is very inexpensive and easily replaced for my purposes on the electric furnace.
I’m not quite ready to apply power to the coils and need to do some preliminary testing of the controller before doing so. The only incandescent light bulb I could find was a 75 watt, so I stuck it in the furnace. In a couple hours it was 160F inside while the shop temperature was 60F. Over night it stabilized at about 185F. There’s probably something to be learned from this about the overall level of insulating performance of the furnace. At this point it’s academic….it is what it is. It’s essentially 2” of K23 IFB plus an inch of wool.
I did some preliminary test on the controller, made few adjustments, and it seems to be working properly. I need to get a little better acquainted with the PID and then I'll apply power to the coils. Meanwhile I think I’ll let the furnace cure for a while at low temp with the light bulb.
– However, I’m getting close to actually being able to melt something now.
....For first melt see post #48 on page 5 here: http://www.alloyavenue.com/vb/showth...uild-Log/page5
Furnace Build Update 10-3-2016
I finished up the electric insert for the larger A60 furnace. It can be installed between the base and main furnace body and intended to be used for lost wax burn outs, heat treating, and other PID controlled temperature use. This uses the same pair of electric coils and controller as the smaller electric furnace and is constructed in a similar fashion. The IFB & refractory work is described at the end of post #6. The sheet metal skin was slip rolled and edge hand formed with a hammer and buck as shown on the smaller furnace above. This electric insert should also come in pretty handy for curing the castable refractory on the recommended ramp/soak schedule. I made a barrel dolly to store the larger furnace assembly when not in use and was thinking I could possibly add a couple guide posts to the dolly to use it separately from the lift and/or as a holding oven.
Furnace Build Update January 8th, 2016
I made the sheet metal skins, cut/fit the ceramic wool, and united all the pieces of my larger furnace.
That center section is an 8 kw electric insert for heat treating, burn out, and controlled heats up to 2200F or so.
I have a 200 watt light bulb in there now and I’ll be interested to see what temperature it stabilizes at. After a few days of the light bulb I’ll use the electric section and ramp and soak PID to run it up 100F/hr.
I’ll turn my attention to adding the lid lifting mechanism for my rig now.
Last edited by kcoffield; 01-08-2017 at 07:52 PM.
Reason: Build updates and picture consolidation
Based upon previous experience I made a rolling carriage to ride on the mast. This along with the counterbalance weight and the block/pulley system are really the heart of the lifting mechanism. I have used similar rolling carriages approaches on other projects on both round and square beams, vertical and horizontal guides, and they are surprisingly accurate and work very well. I had bearings and perches left over from a previous project which also made it an easy decision to go this route. It looks elaborate but it’s just a piece of 3 inch schedule 40 pipe with about 20 minutes of set up and vertical mill work to cut the bearing windows/slots and drill the holes to receive the bosses for the bearing perches.
There is some lathe time into all the little threaded bosses but it’s all just grunt work drilling, tapping and parting off hot rolled carbon steel rod. As a confession, I have an acquaintance that owns a machine shop. I made a weekend visit with some refreshments and loaded a stick of 1”hot rolled into his bar feeder and all the different length drilled and tapped bosses fell off the end before we could get the switch on the machine turned off. I have a little shim punch and just shimmed the bearings to the desired clearance. I made a crude weld fixture to accommodated sliding carriage. Basically just assembled everything, tacked the bosses, removed the bearings, reinstalled the bear perches as weld fixtures, and migged them all up. A similar process was used for the other mast clamps/mounting points. The way it’s loaded, the lower pair of bearings on the furnace side of the mast and the uppers on the opposite side carry the load. The others are pretty much along for the ride but when I shimmed the bearing perches, I set them up with just a little preload. They really grip the post well and are resistant to rotating on the mast, but I included a tracking arm to keep it tracked during actuation.
At the same time, I made the top beam and lower mast clamp that mounts the Furnace Base.
Last edited by kcoffield; 07-19-2016 at 06:43 PM.
Lifting Mechanism (cont)
The lift can accommodate either the 10” bore (up to A20crucible) furnace shell made from the 16” OD drum or you can bolt on a 14” bore furnace (up to A60 crucible) made from a 23” OD 55 gal drum. Both furnace shells have drop-in, replaceable refractory modules. When not on the lift, the furnace shells serve as the storage container for the refractory modules and crucibles and can potentially be operated stand alone as a lift out furnace.
The lift mechanism is counterbalance with the aid of good old gravity. Besides vastly reducing the required actuator power to almost nothing, it allows the actuator to be disengaged and makes it possible to manually lift the furnace when away from power and heated with propane. In theory the effective lifting weight could be balanced to zero. In practice I keep the effective weight of the furnace at about 10-20 lbs without the furnace lid so it naturally seats well on the base. It can be easily lifted with one hand and tuned to about any weight on the lift under 400lbs.
I selected a 30” stroke, 100lb load, 2.8 in/sec linear actuatoras mentioned in the previous lift mechanism post. The lift has an available stroke of up to 44” of clearance above the A-60 crucible. With the actuator directly coupled to the mast carriage it has a maximum speed of ~2.8 in/sec, 30” of stroke, and 100lbf available lifting force. This is how it is configured in the pictures. The pulley system can be configured such that the block/pulley doubles (200lbf) the lifting force while halving the lifting speed and travel or doubles the lifting speed and travel while halving (50lbf) the available lifting force. In the latter configuration, full stroke of 44” is possible if needed though it’s really for greater actuation speed.
The counter balance can also be direct coupled or use a blockand tackle type doubling pulley so half the weight can be used on the counterbalance. The ballast is configured this way in the pictures so applying half the weight of the furnaces provides balance. However, travel is furnace travel is limited to 30” in this configuration because the counterbalance weight must move twice the distance that the furnace is actuated and it runsout of counterbalance track and also because that is full stroke of theactuator. I did make a poor man’s V-Groove guide track so the counterweights don’t flop around. It’s just ½” angle iron stitch welded on the rear guidepost. It works very well and is designed to allow the plate for the counterbalance weights to be easily detached from the trolley if it ever needs to be modified, replaced, or reconfigured.
There were a bunch of great eBay finds I collected over time for the mechanisms. I got each of the v-groove pulleys and wire rope pulleys for ~$1.50 each delivered to my door and also many of electrical components. McMaster Carr was also frequently in play for hardware…love those guys. I place an order at 9:00 AM on a weekday and it usually shows up the next afternoon UPS ground….simply amazing.
Last edited by kcoffield; 04-19-2016 at 10:58 AM.
I wanted the furnaces to be as light as possible so it could be easily moved, lifted, and heat up rapidly. It needed to be made of materials suitable for aluminum melting for electric mode and perhaps up to bronze service for fuel fired versions. Rigidized wool would be the lightest and most insulating but also the least durable. For the refractory module of the electric version, it also needed to have grooves on the ID to support the Kanthal A-1 resistive coils. Typical IFB is about 35lbs/ft3. Some of the better insulating castables are about 70lbs/ft3 while dense castables are 150lbs/ft3. In general, I found the lighter refractory materials to be better insulators but not as strong and durable and vice versa for dense castables. Also, lighter materials generally have lower specific heat so less pounds means less to heat and faster warm up. A 2” wall IFB cylinder would weigh about the same as 1” of insulating castable but I needed to somehow get the resistive coil groove features in there for the electric furnace and I figured making a mold for castable was more work, cost, and risk of fail than just fabricating IFB and wiping on a thin layer of dense castable or Mortar to make it more durable, so that’s what I chose.
Insulating Fire Brick Fab and Assembly for A-20 ElectricRefractory Module
I put a crappy old steel blade on my miter saw and a used upold 80 grit sleeve on the 4” drum of my spindle sander and went to it. Even the dull sanding sleeve went through the IFB like butter. This went very quickly. I took advantage of freakishly warm December (2015) weather here in the US Midwest and moved my saw and spindle sander out in the driveway to avoid making the mess and future health hazard in my shop. When done I just swept up the fines off the driveway and got out the leaf blower….and yes, I certainly did wear respiratory protection, coveralls, and cleaned myself up immediately after doing the IFB work.
Here are the shaped IFB piece parts.
I clamped and glued them up with 3000F mortar. I managed to keep the mortar joints fairly thin, but if I were to do it again, I might thin the mortar just a tad. It was difficult to get enough pressure on the IFB parts with the clamps to extrude the mortar out of the joints and I was a little concerned about putting much pressure on the fragile IFB but it turned out pretty good.
After the mortar set up over night, it was back to the spindle sander and few simple OD/ID jigs to true things up to final shape and dimension.
Then I glued the subassemblies together.
Now things get a bit more interesting. I designed the coils to have up to the maximum recommended wall and coil loading. I iterated a bit and came up with two resistive coils that I’ll elaborate on in the Electrical Controls section. The result fit very nicely into two coils that made three circumferential trips each around the furnace ID. This had the added benefit ofbeing able to run one or both coils at once. So I had to figure out how to cut a double lead groove on the furnace ID for the two coil grooves.
These coil support grooves are nested 180 degrees out of phase to each another, in a 3” pitch, double lead, spiral groove, on the 10”furnace ID, so the two grooves nest inside one another like a double lead thread. Each coil takes three circumferences to travel bottom to top of furnace wall with the intent of promoting uniform wall loading when operated together or independently. It may take some tweaking to get the heat distributed evenly. Each groove is ½” wide and 5/8”deep. The coils will then be attached with wire through the wall in the grooves for coil retention.
I had all kinds of elaborate thoughts about how to cut the groove in the IFB and in the end the simplest approach won out……cardboard!
A flat-layout cardboard template was made for the double lead groove at the 3”lead and furnace ID along the active length of the furnace wall. Slots cut the straights with a utility knife. Then the thickness of the cardboard guide ribs were built up by laminating two layers of ¾” wide by 1/8”thick cardboard strips bonded with spray on contact cement such that the guide grooves were nearly ¼” deep. This basically made a ¼”+ wall, 10” ID cardboard barrel with the required double lead spiral guide grooves. It was remarkably stiff, gripped the abrasive IFB ID surface well, and best of all price…was free.
A shoe was made for my router base to conform to the furnace ID with ¾” diameter guide bushing. The burr just cut through the cardboard support webs on the first layer of the template as it went along as it didn’t matter if the previous guide groove in the template (behind the groove that had already been cut) was no longer supported and held in place. However, this did require the entire groove be cut in one pass. This is where a thin mortar joint was important. A $7 round nose ½” diameter double cut carbide burr gave its life to complete this task. The IFB was very easily cut but even the thin mortar joint was very hard and I didn’t know if I’d get through all the joints (6 trips around 10 IFB segments was 60 mortar joints in all!) before the carbide burr gave up the ghost. Remarkably, after cutting 60 mortar joints and about 18ft of IFB it still looked like a burr but I suspect it won’t be much good for cutting metal.
Happy to report it all worked like a champ and I have a very uniform ½” wide, 5/8” deep, round bottomed, pair of coil support grooves!! – I was very relieved. It had the potential to be an epic fail. If you look close, you can see there were just two very minor chip outs in the groove where the mortar blew up on me a little, on two of the especially stubborn-to-cut mortar joints. These can be easily touched up when I wipe on some hard face. This (groove cutting) was the highest risk part of the IFB refractory build. The pucker factor was quite high so very glad to get through it ok.
Here are the IFB assemblies with furnace lid. I used same approach with IFB and will add a thin coat of dense castable. I did add a little taper to the top of the furnace and the lid so the lid could more easily land in the upper few inches of the barrel.
Making the lid slightly larger than the bore as opposed to the entire furnace OD kept it smaller and lighter and easy to lift by hand. I cut some grooves into the IFB and filled them with dense castable for extra strength and then gave it a thin hot face.
I decided since the lid was small I’d just make a different one that had the proper vent size for the fuel fired version of the furnace.The small hole in the present lid is sized to fit the OD of my thermocouple sheath and degassing lance.
I’ll add a couple photos after applying hot face by painting on some mortar and some dense castable.
All of this took sixteen K-23 IFBs 9”x 4.5”x 3” (2 boxes of10), ten 9”x 4.5”x 2.5” (box of 12), and a small part of a pail of mortar. I had 8 hours into making the IFB structures and then an additional 4 hours into making the template and cutting the heating element grooves. It was about $3/brick. I’m lucky to have an acquaintance that owns a full line refractory supply business.
At the time of this post the electric furnace module is waiting for coating, addition of the heating elements, and outer skin. –Stay tuned for those future additions to this section.
A60 Castable Refractory Module
I have not yet started the larger castable refractory module but have some ideas for it's construction. Watch this space for future updates.
Furnace Build Update May 15, 2016
Since I had plenty of it left over, I diluted some of 3000F mortar with water to a consistency that filled the IFB pores, brushed, and leveled well. It dries very fast as the IFB wicks moisture from the coating. It really increases the surface strength of the IFB, makes a significant improvement in durability, and makes the IFB sort of look and feel like a castable refractory. It noticeably adds some weight too because with even in a thin coat, the mortar is very dense. Hopefully it will stay put through heat cycles. I also scraped in some drain channels to the furnace base and made a plinth pedestal to center and position the crucible. The pedestal (not shown) is mortar coated IFB. I plan on making a plinth cap that is dense castable to make it easy to place the crucible centered in the furnace so hopefully I’ll never set the furnace body down on my crucible.
Furnace Build Update September 23rd 2016
I built the refractory module for the electric insert for the larger furnace. This one has 14” bore, 22” OD, and 7 ½” tall whereas the unit above is 10” Bore. Same method of construction as used on the electric furnace module above. This uses the same pair of resistive Kanthal A-1 coils (only 1 1/2 circumferences for each coil instead of 3) and controller as the smaller furnace and is intended for use in heat treating, lost wax burn outs, and other controlled temperature use with the bigger furnace. However, I intend to use LP or NG fuel for aluminum melting with the larger furnace.
Furnace Build Update 12-26-2016.
It may be routine for many here on AA but the design, selection, and use of refractory materials was new and a learning opportunity for me for my furnace build. Cutting and piecing together IFB with mortar was fairly straight forward but working with castable refractory was new ground. I’ve used many materials and net shape forming processes in machine building and thus my interest in metal casting. I thought other net shape processes may be applicable to the placement of castable refractory materials as well. The most similar processes seem to be molding concrete and ceramic statuary and figurines though castable is clearly a different animal than concrete and it seems not all castables behave the same.
So for my larger A60 furnace I decided to try some different things in regard to the construction of the refractory linings. I used a combination of insulating anddense castable which certainly has been done many times, but with a few variations. The result is a composite structure consisting of both insulating and dense castable, molded in several parts and successive steps.
This approach does of course place strongest most refractory material (dense castable) on the interior of the furnace where the highest heat will be encountered and then places the higher insulating value material (insulating castable and then wool) in successive layers outward. It also tends to minimize the entire mass of the castable refractory which is helpful in managing its overall weight on my lifting mechanism and the lower thermal mass should aid in it heating up more quickly, all of which were objectives of the build. There’s some modest economic benefit as it doesn’t use as much material and since I only had two bags each of dense and insulating castable to work with I wanted to plan accordingly and see what I could manage within that amount of material. There is additional benefit in that it allows mixing and casting smaller batches for each piece which was helpful in allowing me to stay well within the recommended working time for placing the castable. Also, if you should happen to have a failure, you don’t lose a large amount of material.
Besides mold complexity, the tradeoff may be somewhat less durability than more massive refractory structures but being used in a lift-off furnace, mechanical abuse should be non-existent and the primary issue becomes what it is for everything……………. how well it survives heat cycles. There may also be some unknowns in expansion rate of the two materials but this is always the case with structures even those made from homogenous material because during use there are large temperature gradients across the structure and this induces large thermal stresses which can lead to cracks. The toughest thing on refractory structures is heating and cooling and the more rapid the heat cycle more stressful it is. Only time will tell on this factor.
Similar to the previous experimentation with the molded dense castable plinth in my previous post (post #68 on page 7), and lessons learned there, I made a mold out of construction lumber and mummified the pieces in packing tape to serve as release surface. For those interested, I just rough sanded the surface of the wood during fabrication on my spindle and belt sander and then apply one generous coat of shellac to seal the wood. I then briefly hit each piece with a 220 grit foam sanding block which provides a surface the tape will more aggressively stick to than just raw wood. This is much quicker than trying to work the wood surface to a finish that could be released from castable refractory. I can also be quickly and easily repaired.
The first molded part was the flange. The typical wall thickness of the flange is.75” before draft. I have about 6 hours into making the flange mold and thoughit may look like a lot of work, the difference in invested time isn’t much different than building the simple mold for a heavier structure. I used this flange mold three times to make two identical flanges for the main furnace body and one for the furnace base. It takes about 20 minutes to de-mold a cast flange,30 minutes to clean, re-assemble, and prep the mold for another casting, and another 20 minutes to mix and cast another part, so the multiple uses recovers a bit of the time invested in making the mold. Overall the flange is 22” OD, 15”ID, 3” tall on the inner ID and 1.5” tall on the out OD. All lateral surfaces have 5 degrees of draft except those on the inner/outer diameters where the sheet metal mold hoops I used have no draft. The hoops are also covered with packing tape, are split, and taped together at the joint. To de-mold, I remove the bottom plate, run sheet metal screws into the back side of each of the wooden segments and pull them with a claw hammer from the drafted side of the mold. Then I remove the wooden structure so only the metal hoops remain. After cutting the taped joint on the hoops they peel right off.
For insulating castable I used Harbison Walker Greenlite 45L (GL45L). It’s supposedto have good strength to weight ratio and insulating properties is 2500F/1370Crated. When cured its density is 71lbs/ft3 which is about half the density of dense castables. The Greenlite product is strange stuff and worked much differently than the dense castable. It calls for 23% water by weight for vibration and 24% for hand packing. In contrast the dense castable I used was Harbison Walker Ultra Express 70 (UE70) which is 157 lbs/ft3 and recommends only 6.8% water by weight seemed unworkably dry. I had to mix it slightly wetter than that to make it respond to vibration and be castable. The UE70 is hard, strong, ceramic fiber reinforced, and supposed to cast well into detailed structures. It’s rated for 3000F/1650C duty and continuous aluminum contact. The Greenlite 45L was much easier to mix of the two, but both responded and flowed very well under vibration. For the GL45L, if I dropped small dab on the mold fixture it would just sit there. As soon as vibration was applied it densely sucked together like a bead of water on a waxed surface, so much so, it was difficult to stick a 3/8” dowel into the settled mixture in the deeper portions of the mold shortly after applying vibration.
The Greenlite filler material was light and seemed to be about the size and consistency of walnut shell blasting media. The filler is small and fairly uniform in shapeand size between .030-.060” in diameter and seemed to act almost like marbles in the refractory cement perhaps possibly explaining why it flowed so well and densely self packed into the mold detail under vibration.
Ic alculated the volume of the flange mold and estimated the total weight of insulating castable based upon 71lb/ft3. I weighed the castable and water using an electronic scale for the castable and triple beam scale for the water. I mixed it in two separate successive batches for each flange.
I leveled the mold, turned on the vibrator, and started scooping in the mixed refractory at a uniform depth around the mold and watched it flow and level out. The first batch filled all the lower features of the mold, basically everything except the top annular ring. I then mixed the second batch and finished it up and applied vibration again to the entire mold. I put a couple 100 watt light bulbs in the furnace barrel under the base of the mold and left it overnight. The next day itwas solid and well cured. It has a few bug holes but in all, it’s a solid usable part, and given the shape complexity, I was satisfied with the result for a first pass.
Here is the mold and the resulting insulating refractory flange casting.
On the second flange, I still mixed two separate successive batches of castable but also made a small wand vibrator to work the deeper features of the mold but it didn’t seem to get a better result than the using just the entire mold vibrator and occasional tucking with a dowel rod (which was probably also unnecessary) so I didn’t use the wand. I only vibrated the mold for about 5 minutes this time. It had about the same population of bug holes previous part buthad nice tight density at the open/exposed mold surface.
On the furnace base flange I made an insert that replaces one of the eight mold segments to incorporate the Tuyere tunnel feature. I made the Tuyere plug out of pink polystyrene insulation board and wrapped it in packing tape. After coating each end face of the Tuyere plug with shellac so the polysester solvent would not attack the polystyrene, I stuck the foam Tuyere plug to the steel hoops with a thin coat of bondo (polyester auto-body filler) and stuck a screw through the steel hoops into a well of wet bondo that I had ground into each face of the foam plug. This provided an added measure of holding power as the plug would tend to float in the wet castable if became dislodged during vibration. To de-mold the foam Tuyere plug I originally thought I’d just pour acetone on it to dissolve the plug but instead I just dug center out with a chisel so it would readily collapse and pulled the rest out by grabbing the packing tape with a pair of pliers thus avoiding having to cleanup the gooey mess.
The lost foam approach worked great for forming the Tuyere feature. If you use polystyrene foam with castable refractory, you should seal it with shellac or something to form a water barrier that doesn’t attack the foam as the foam can wick water away from the surface of the castable refractory and affect its curing and/or affect its cured mechanical properties. The packing tape works very well in this regard but if you have a complex surface and wanted a very smooth wrinkle-free finish, shellac or other brush on coatings may be a better choice. It just needs to be something that seals but doesn’t attack the foam such as alcohol suspended shellac, epoxy, or latex paints. Coatings reduced with lacquer thinner, mineral spirits, acetones andother strong solvents will melt polystyrene. I don’t think polystyrene would bedurable enough for ramming but it seemed just fine for service in vibrated castable.
….and here is that molded flange with Tuyere feature. The Tuyere was long enough that I couldn’t get the entire feature by replacing just one mold segment and I didn’t want to make a second Tuyere segment so I just doctored up the rest of the feature with a small patch after I de-molded the part.
With the cast flanges in hand, I used them as part of the fixture to cast in place the dense castable thin-walled hot face. The furnace body hot face is cylindrical, 14” ID, 19” tall, only 5/8” wall thickness between the flanges and ½” wall at the flange interface. The dense castable hot face is cast and bonded directly onto the insulating castable flanges and the slight wall thickness change provides some mechanical retention features cast in as well. Each flange weighs about 12lbs and the hot face cylinder about 54lbs.
Here’sthe finished main furnace body. 14” Bore, 19” tall, 22" Flange OD. Total weight of body is 77lbs. 12lbs each flange and 53lbs for cylindrical hot face which is only 5/8" wall.
Similar to above, I used the insulating castable flange with the Tuyere feature as part of the mold to cast the composite furnace base using a combination of dense and insulating castable refractory. Since I was only making one piece, I used an expendable lost foam Tuyere and foam plug to form the basin. Here’s some build photos and the completed composite refractory insert for the furnace base before and after de-molding. The photo of the bottom side sort of looks like an Oreo Cookie landed on a flying saucer.
So by now, I have everything except the furnace lid made. I have some additional things I’d like to try on the lid but I’m out of refractory so that will need to wait for a while. I’ll post after the next session.
Furnace Build Update 1-2-2017.
Having gained some confidence with the previous castable refractory parts, for the furnacelid I decided to make a much more challenging shape. Now without a doubt, a simple disc with a hole in the center would have worked fine but I wanted to explore the art of the possible for future use……and well, just for the heck of it. The lost foam plug I used for the furnace base basin and Tuyere tunnel were so easy to form and worked so well I decided to use lost foam more extensively as the mold material for the lid. So I made some wooden structure and hoops for the cylindrical features and went crazy with the lost foam aided shapes. -I sure do like my over-arm router. Best shop machine I ever made and sure makes this type of pattern work go fast.
As before, it was poured in several steps, dense castable first and then insulating castable on top, all using full mold body vibration to flow and settle the castable refractory material. The actual casting was done in three pours. First was the dense castable hot face layer which was cast about ½” thickness everywhere. The dense castable was poured in two steps. First the lower portion was cast and allowed to set up well enough so a temporary mold could be used to cast the upper vent hot face cylinder. After a couple of hours that had set up well enough that I could remove the temporary vent mold, install the six cavity segments and cast the insulating casting at 5/8-3/4” average thickness.
Here is the De-Molding and why they call it LOST FOAM!. The particle board hoop support structures are all split so they remove easily without stressing the still green part. I just put a burr on my die grinder and hollowed out foam plugs enough that they could collapse on themselves and pulled the remainder out by the tape skin. The burr made short work of foam and it went quickly albeit a bit messy. I had put a band of filament tape (which is much stronger) under the packing tape around each segment and after hollowing, I grabbed the filament tape with pliers and pulled the remainder of each plug out. De-molding went very quickly.
So the lid has ribbed structure on top and bottom with the bottom ribs able to channel air flow with the crucible raised on the plinth or flow conventionally through the vent hole with the crucible place deeper in the furnace. I don’t know if it will work any better but probably won’t be any worse than a flat lid with a hole. I had a whole day available in the shop and it was mostly just me experimenting and jacking around to see how far I could take the various molding methods working with castable.
Here’s how it came out. It’s a pretty crazy shape. The foam hollow segments on top and around the periphery save about 13lbs of weight/refractory and will get stuffed with ceramic wool and a sheet metal cap and ring around the circumference for lifting mechanism attachment points.
Here it is on the furnace body.
This and the remainder of the refractory castings will get a very slow cure. This post is repeated on page 8, post #76 with bigger pictures. -That's all for now.
Last edited by kcoffield; 01-02-2017 at 10:48 PM.
Reason: Build Update & Picture Consolidation
Electric Furnace Power Management/Controller
At these power levels and for safety, I spent some money on the components for the controller. I made the controller with the intent of being able to us it on this furnace or a larger electric burn out or heat treat oven. It has an illuminated, dual action on/off main power switch that energizes main power to the panel through a master contactor and an emergency stop that disconnects all power. It has a PID that directly controls two SSRs, one foreach furnace coil. To protect the coils, each is fused with a 25 amp dual pole circuit breaker. Each SSR/Coil can be isolated with a toggle switch so they can be operate separately (4.1kw maintain mode to free up a little shop power) or together (8.2kw). Each coil toggle switch just opens the low voltage control circuit to each respective SSR. There is a status light and hour meter for each coil. For grins, I bought a $7 cheapo power meter from eBay/China which displays voltage,current, power and energy. It has an inductive current sensor.
There’s a K-Type thermocouple plug in the bottom of the panel connected to the PID. Either furnace temperature or molten metal temperature can be sensed. We’ll see if the K-Types hold up. I bought three Mullite sheaths ¾” diameter by 24” long off eBay for $15 each. One is for molten metal immersion, I intend to grind a small hole in the end of to use as a degassing lance, and the third is a spare. The sheathed thermocouple mounts above the crucible and is manually lowered into the metal for molten liquid sensing to hold/maintain pour temperature after melt has occurred. I got some great deals but spent some money on the electrical controls and conductor. With the separate circuits and all the bells and whistles it took me the better part of a day to wire it all but it came out well and will hopefully provide reliable service, good control, and melt/operational status information.
Linear Actuator Controller
The whole furnace is counterbalance so it doesn’t really need an actuator but it’s a case of needs versus wants I guess. I had a 12v DC power supply and control components leftover from an acme screw linear actuator I made for my over arm router years ago. I had also made a remote foot pedal station for that project. I made the patterns and castings for the base and pedals years ago and have spares that looked to be a good candidate for the job. This was basically a ready to go foot control and power supply and insteadof a lead screw and nut it would just drive the packaged linear actuator. I actually decided not to use the foot switch figuring it and the cords would just be something to get in the way and be tripped over and instead opted to make a mounted control panel and remote control.
It was $15 eBay score for a remote and panel control that included a control logic board. I would have had more than that wrapped up in the relays and discrete components it replaced. The wireless remote pretty much came along for free so I decided to give it a try. The logic board allows selectable control for either momentary or latching so you can activate for full actuation stroke against the limit switches or momentarily bump it along to desired height and this can be done with either remote or panel mounted rocker switch. This simplified the wiring and components quite a bit as it only needed power in, power out to actuator, one common for all limits switches and one wire to the normally open side of each limit switch. I hardened it up a bit but the real question will be durability.
I added a speed control for the actuator (no kidding $4.95on eBay so I bought 3). I was going to use a staged/intermediate limit switch forthe down stroke that greatly reduced the speed about an inch before the lifted portion of the furnace contacts the base so it lands softly. It wasn’t needed. I found the speed of the actuator when direct coupled to the mast carriage was fine and since the actuator coasts a bit after power down, the limit switch setting is adequate. The staged limit switch is there if ever needed. I bent up a box that package up well with the rest of my build.
I made a wiring harness, routed it inside the guide post, and popped a few holes in the post to mount receptacles and strain reliefs for the cable. It connects the limit switches and actuator power plugs to the actuator control box via a harness and connector. I made the harness from left over wire and sleeving from an electronic fuel injection conversion I did for my car. The plugs are audio plugs left over from the actuator cabling on my over arm router build some years ago. This keeps things pretty clean and protected. I can also unplug and completely remove the actuator in about a minute for manual actuation of the furnace lift. I made the limit switch boxes from 1/8” scrap.The lower box can be moved but it doesn’t need to during regular use so it’s fixed to the guide post with a couple screws. The upper limit switch can be re-positioned. With the actuator directly connected the upper switch isn’t actually needed because the actuator has an internal limit switch a full retraction and extension. The electric furnace controller can also be easily removed for all manual/gas only furnace operation but for my use will just stayput and come along for the ride.
Resistive Heating Elements:
I used the Kanthal design handbook to back my way into thepower level of the coils designing to be at or below maximum recommended furnace wall and coil surface loading of 3.3kw/ft2 and 26 w/in2 respectively. There are two, 14 gauge Kanthal A-1 coils, wired in parallel.
So this calculated to requiring two coils 756” inches (63ft) long wound on ½ OD coils, for 13.46 Ohms each (resistance at 1000c). These are wound in ~½” OD coil diameter. At 0.0641 wire diameter this results in a 30.8”close-wound coil length so the coil requires a little more than 3:1 coilstretch for three trips around the 10” ID furnace. This requires about 1.2 lbs of 14 awg Kanthal A-1 (~1lb/100ft for 14 awg) for both coils. Kanthal A-1 isabout ~$40/lb. This equates to a power requirement of 8.2 kw.
So all of the above should produce a result that approaches the maximum recommended power density for a Kanthal A-1, coiled resistive element, for an electric furnace of this size (can accommodate up to an A-20crucible).
14 awg gets a little stiff to wind that tight by hand so I made a coil winder that runs off a cordless drill.
I had a piece 3/8-18 threaded rod which I thought would workout pretty nice since the wire diameter is the same as the pitch of the thread so the wire lays into the root of the thread as the rod winds and advances. In practice, a smooth rod would have been better because the thread and the wire diameter were so close (wire diameter was actually .001 more) it bound up after a few inches. The small groove that feed the wire and the drill bit with split bushing to clamp the wire worked well. I added a tensioner which was just aspring loaded brass screw that bore against the wire. It seemed to help but I probably could have just held the wire to maintain some tension. I made an additional pair of coils for the electric insert for the larger furnace and used a smooth 3/8" piece of tubing and it self advanced and worked very nicely producing two perfectly close wound coils that wound in about 30 seconds each with the aid of a cordless drill.
I’ll use the twisted pair method to reduce the resistance and temperature of the feed through wire. The coils are held in their grooves with Nichrome wire ties.
Most Kilns seem to immediately route the wire outside the insulated area to allow low operating temperatures for the conductor. One disadvantage to the pair of spiral elements arrangement is the connection to each coil is at top and bottom of the barrel and each element’s connections are on the opposite sides. I wanted to be able to use the tuyere as the electrical connection exit. I figure the temperature near the out diameter of insulation will be between 200-400F. After some investigation I discovered MG and MGT furnace wire. This stuff is great and was new to me. It is rated for operating temperature up to 450C UL and 1000F non-UL. It’s nickel plated copper to prevent oxidation, very conductive (high ampacity), and can carry much more current than ordinary wire for a given gauge at room temperature. The 18ga is rated at 31 amps @86F and 17amps at 600F. It will be located vey near the outer diameter of insulation and will probably actually experience about 200F as furnace is operating. I used 16 ga for a little extra cushion as it was essentially the same price to do the job…$10.
Watch this space for further updates on installation the installation of the resistive coils and completion of the electric furnace.
Furnace Build Update May 15, 2016
I finished up the Kanthal resistive coils and prepared them for installation by doubling the leads and stretching them to length.
When I stretched the coils they recoiled quite a bit and I took a look at them before stretching them to full length. I’m glad I paused to look because the coils were not stretching evenly along the length of the coil. Not sure why but I massaged some the tighter sections a little to make them uniformly spaced before giving them the stretch to final length. On advice from Sandcrab in post #34 (thank you SC), I over stretched the coils a bit with the intent of putting them into a little compression as they were installed in the grooves….and mostly succeeded. There were a few sections that weren’t quite fully seated in the grooves. The second set of coils I made stretched very evenly so I suspect uniformity in coil diameter and spacing is key.
I made (32 of!) these retainers out some leftover Kanthalwire and stainless welding wire, wound up these little gripper springs, and installed them in the grooves every 90 degrees. If I ever need to replace a coil I’ll just snip and replace the retainers.
I bound the MGT wire to the Kanthal leads with some steel lugs I made from ¼-20 UNC coupling nuts. I’m not certain what the temperature will be at the MGT/Kanthal junction but am expecting mid hundreds Fahrenheit orless. I had cemented in some rivet nuts with mortar to the IFB furnace body to secure the MGT wire, installed 1” of ceramic wool, and finished routing the MGT wire near the furnace exterior and to the exit location on the IFB. The wire exit location lines up with the Tuyere tube on the furnace shell.
Here are the coils installed:
Last edited by kcoffield; 10-05-2016 at 11:25 AM.
Accessories and Tools
Where would be without ingot trays? I salvaged angle iron shorts and drop from the bin and cut and welded them together to suit my crucibles. My friends asked me what use I had for short 8” pieces of angle? Good enough for the small ingot trays. They store in the mobile base. The small ones can be accessed by rotating the cabinets. The larger ones slide onto/off of the shelf on the mobile base from the bottom or the furnace base clamp can be loosened on the post and rotated out of the way. The different lengths suit my three crucibles.
Pour Carting and Shank
First Order of business after or near furnace completion is something to handle the crucibles with. The plan at the moment is a pouring cart/trolley with interchangeable jaws to fit A10, A20, and A60 Crucibles. This way it’s easy to add a set of jaws if I get different crucibles. Only Sketch so far but based on another pour cart recently discussed here on this forum. It will have a support tube so the shank is adjustable along with the furnace base (crucible) height to accommodate different mold heights.
Also early on my to-make list are skimming spoon, tongs, degassing lance, pyrometer. -More to come in future Posts
June 13, 2016 Edit/Addition of Pouring Cart Progress
When I started thinking about building a pouring cart and shank, something with the capacity to handle the A60 crucible was the most challenging so I made that my starting point. My original intent was to build a single clamping mechanism with interchangeable jaws to handle all crucibles but when I was doing the layout work, the A60 is such a beast even the A20 gets lost in it.
This is the progress and seems to be going as planned so far, at least for the A60.
Though it would have been easy just to build an extra set of smaller jaws for the other crucibles the outer diameter of the A60 clamp seems impractically large for the smaller crucibles and may potentially cause interference with molds and large pouring distances. So, the A60 may need its own clamping mechanism. Since it’s small and light it’s hard to beat just an open ring shank for the A10 but at 35-40 pounds all up, the A20 is sort of a tweener. I may build a smaller clamp for it or revisit an open ring shank. Think I’m going to need to finish the larger one and experiment a little pouring sand with it to see how I like it before proceeding with any adaptations for the other crucibles.
July 10th 2016 Edit/Update. Pouring Cart and Shanks
So, I decided the A60 gets its own active clamping mechanism for use on the pouring cart, the A20 an open ring shank that can be used on the pouring cart or free hand (though it’s a bit of a handful), and the A10 just free hand open ring shank.
Here’s a couple pictures of the A60 on the pouring cart. I had to stoop over quite a bit with just straight handle bars so I made myself a couple of raised handle bars (ape hangers for you Harley guys). They are adjustable and can be mounted ahead or behind the over center clamp. There’s a lot going on with the cart but I think it will be a good tool and adds a measure of safety and control to handling a lot of molten metal and……I had fun making it.
And here’s a short video of snatching and replacing theA60
I have put 50lbs of weight in the A60 on the cart and also filled it with sand and made some practice pours and can maneuver it well. Same with the A20. As I mentioned earlier, the A20 is a bit of a handful free hand with molten aluminum weight. I still may add provisions for counterweights on the tail shaft, and maybe a kickstand to park it. I considered making a tricycle out of it. Although my furnace would allow an extended caster wheel under the crucible, it would interfere with molds that sat on the ground. The third wheel could extend backward but being able to teeter the cart makes it more versatile and maneuverable. So for now I have omitted any third wheel.
I would like to send a thank you to Junkyard. His furnace and comments about open ring shanks at the thread below were helpful and caused me to revisit their use for A20 and A10.
In retrospect, I believe I could have made an open ring shank for the A60 as well similar to the approach I took on the A20. However, I am happy to have the active clamp and jaws. It seems to work well and will be easily adaptable to other large crucibles if ever needed.
Here’s the open rings for the A10 & A20. Since I already had the mechanical actuation on the cart I added the top clamp to the A20 to further secure the crucible. The A-10 seems fine without it. I show a removable 90 degree handle on the A10 shank but in use I find it more comfortable without just using the shank handle.
…and a short video of the A20 on the cart.
Since I can adjust the height of my furnace base and thus crucible height in the furnace, I can just set the furnace and cart height to suit the pour height, so I don’t think I’d necessarily need to be able to change the pour cart height on the fly unless pouring multiple significantly different mold heights in one melt. I thought since the A20 is just a static open ring and needs to be lifted onto the crucible, I may need to be able to adjust the pour cart height on the fly but with a little experimentation I discovered I can do this satisfactorily by just tilting the cart as shown above in the video.
Because of the lower initial height I can only get about 8”of active vertical travel between the two telescoping tubes on the mast of the pour cart. I have a pneumatic air cylinder left over from a previous project that would do the job but I think I'll buy a linear actuator and small sealed 12v battery that quickly installs with two hitch pins to the provisional mounting points on the cart. It could also serve double duty as the X-Y table lift on mydrill press.
Here’s a few build progress photos.
First the over center clamp. It actuates an inner rod. I later added a safety latch to the clamp handle as seen in the earlier photos.
Here’s the cart axle and mast. I used a pair of long 5/8”shoulder bolts for the axles.
I needed some larger diameter steel wheels for the cart so If fabbed up a pair. They are ~12” in diameter. Each one has 18 pieces in them! I got a little carried away but had fun making them. The 1/8” hot rolled hoops were too thick for my slip roll so I just hand formed them around a chunk of 6”pipe and tuned them up with a little hammering and then clamped and welded them to the rims. I stitch welded them sparingly, and cooled them between every few welds to keep them flat and true. I also added bronze bushings to ride on the shoulder bolt axles.
To make the rings conform to the crucibles, I measured the diameter where I wanted the rings to land by wrapping a piece of paper around the crucible. I fed a piece of ¼” OD tube through my slip roll as a test piece. Then I calculated the conical layout radii at the two diameters, cut a test piece from a scrap of 18 ga, and rolled it up. Then I cut the 1/8” thick conical piece which I hand formed and hammered around the 6” pipe similar to the wheel. I stitch welded them sparingly and let them cool between every few welds. They stayed put pretty well and only needed minimal communing with the hammer afterward.
So, I can now handle a hot crucible………getting closer to meaningful casting.
Build Update July 24th, 2016
I added the linear actuator to the pouring cart to enable changing height on the fly. Not sure how much I will use it on the pouring cart but it may come in handy for pouring multiple/different height molds. The switch, cable, battery, and actuator can be quickly removed and repurposed as a lift for my drill press x-y table where it’s likely to see more frequent use.
The actuator is powered by a small sealed lead acid battery which is good for more than a hundred cycles before recharging. I built a mechanical interlocking rocker switch that incorporates two double pole, single throw, normally open, momentary switches to provide the up/down controls. I added a slip on dust cover for the switch. The actuator is similar to the one on my furnace lift but is rated at 400lbs load and 1”/sec rate of lift.
Here’s a short video of the actuator.
Last edited by kcoffield; 07-26-2016 at 03:12 AM.
wow, thats a hardcore build right there. awesome build.