How to best measure molten aluminum temperature within a crucible?

I am in the initial design-stages of building an electric furnace - for casting aluminum.

Question - what is the best way to measure the temperature of the molten aluminum within the crucible?

Discussion - I've seen videos of people using a laser-thermometer (shot at the top the aluminum) to obtain temp. Is this an accurate measurement? The only downside I see with this method is that I'd have to expose the top of the crucible to obtain a temperature - dumping heat every time I lift the lid.

Is there a direct insertion thermometer (or thermocouple) intended to be inserted into molten aluminum?

Better question - how do others (using electric resistive-heating furnaces) verify that their aluminum is at proper casting temperature?

 

Kelly (Al2O3) has an electric furnace build thread on this forum and has a removable refractory plug in the lid to allow a thermocouple to be inserted into the melt without opening the furnace as well as for an argon gas lance. A cheap stainless steel sheathed K type thermocouple is up to the task but will need either a carbon or ceramic sheath to prevent the stainless steel eventually dissolving into the aluminium melt and contaminating it.

 

I am using a budget setup as Mark described: 25 $US HF multimeter with a K-type input, 15$ 4"/100mm K-type probe with a homemade 2$ graphite sleeve.
I have a Mifco thermocouple in 1/2" EMT conduit waiting in the wings for me re-wire the output end to a yellow K plug.

 

Very cool. Use an inexpensive carbon-arc rod, drilled down the center – to shield the thermocouple from the molten aluminum. Carbon-arc rod is available from small, to rather large sizes. Is there a point of diminishing returns (regarding thickness of the carbon that's shielding the thermocouple)?

Follow up question – for a PID-controlled electric furnace (SSR), presumably I would use a permanently installed Inconel thermocouple to measure/control oven temperature, and the carbon-shielded thermocouple would be used only to spot-verify my pour temperature. What would be a typical set-point for the oven temperature? More specifically, assuming my target aluminum pouring temperature is 1300 degrees F – would it be reasonable to set the oven to 1600? I suppose there’s a balancing act between short melt times and overheating the aluminum.

My goal is to ultimately cast some jet boat exhaust manifolds for a Pontiac 455 engine (via lost foam casting). I have a set of Olds 455 manifolds that I am going to cut open - to inspect how the water jackets were originally patterned on the inside. Each manifold is cast in two parts - the main manifold (about 12 pounds), and the manifold-riser (about 10 pounds). Pretty straightforward. The internal coolant passages however are a mystery (until I cut them open). I'm not sure how much "contact" is required between the inside and outside of the water jackets. I'm also interested to see how they address thermal expansion issues within the water jackets.

For Kelly specifically, I've been watching your youtube videos (in one of your videos you mentioned this website, which brought me here). I apricate you taking the time to make & post your videos. In your 8 kW furnace build, I noticed that you targeted an energy-flux of 30 watts/in^2 for your heating elements (14 gauge Kanthal A1). You are within the recommended range of 20 and 60 watts per in^2. The larger gauge wire will take more punishment (higher max temp). I'm assuming that being on the lower end of the energy flux is beneficial (ultimately extending the service-life of the wire)? Today I made up an excel spreadsheet - to assist with designing heating elements from Kanthal A1 wire (calculating number of turns based on mandrel diameter, overall length of element once stretched, resistance, energy flux, etc). I don't want to copy your design, but I also don't want to reinvent the wheel either. Question - how did you settle on the 8 kW power input? And also the 30 watts/in^2 energy flux? I'm trying to size my furnace, and design my required heating elements, but I'm not comfortable making a final decision (yet). I just placed an Edelbrock Caddy 500 intake on the scale - it's 20 pounds. Honestly that's probably the largest thing I would desire to cast in one pour (or perhaps an aluminum bellhousing - which would probably also be about 20 pounds). So does this put me into the 8 to 10 kW range, with an A60 crucible?

I'm also trying to find a local source for some A356 casting ingots (in the Seattle/Tacoma area). I found a few places back east that will sell it for as low as $2.30 a pound, but the shipping fees practically double the cost.

 

Eric, you live near this guy

https://www.ebay.com/itm/1850798530...JPcD30jI7RqD0eDWwjNW1f2Bkgin|tkp:BFBMtv3imJli

Drive over and pick it up. I have bought several times—-good stuff. Fair price. Make a no-shipping deal.

Denis

 

Thanks for the link Denis. It looks like this guy is melting down old cylinder heads and engine blocks. Question - wouldn't it be much easier to melt down aluminum rims?

I've read that cylinder heads are some of the best casting material for the home hobbyist. Not exactly sure why. It's a lot of labor to completely prep the head for melting - pulling it from the engine, disassembling it - also removing the pressed-in components (valve seats, valve guides, blind-steel studs, etc). And for aluminum engine blocks, pulling cylinder sleeves is no easy task either. Some engines the sleeves can only be machined out. One would think that aluminum rims would be the go-to source for making casting ingots from recycled scrap. Simply pop off a tire, media-blast the paint/finish (if present) then melt away. If both engine castings and aluminum wheels are typically made from A356, then why bother messing with engine parts?

Also curious. I'm looking at the pictures of his furnace. What's actually holding the molten aluminum? It looks like his furnace is lined with loose ceramic fiber. If he's using a steel pan, are there any concerns with iron contamination within the ingots?

I guess a better question would be - is it best to purchase new/virgin A356 casting ingots, or to make (or purchase) them from recycled car parts?

 

You'dhave to ask him He's been at it a while.

Blocks are more compact I would think and don't require breaking.

I think it lined with a trowlable refractory. He is using a muffle furnace I believe. Art won't claim his ingot are true 356, but they are darn close. Good enough for most of us I think.

I've been happy. Icould get rims and melt them, but it is a matter of time. The boxes come in conveniently sized ingots. Put em in the furnace and flip the switch.

Denis

 

Eric
I just got my last batch of aluminum from a new charitable supplier (read "free"). It's a local automotive performance shop. The kind of place where you take your engine, not your car. They supplied me with four Chevy heads. They were pretty much free of the kind of grime you'd get from tyranny cases (my former favorite) and thus a cleaner melt. Greater mass too. You made a true enough observation about the studs, but mine turned out on these heads this time. It took some elbow grease, but not too bad. I cut the heads up with a horizontal band saw and spaced the cuts to go approximately down the centerline of the valve seats and valve guides, so they just mostly fell out. I accessed the other steel components like the core-hole inserts, etc with the vertical bandsaw. In a few cases i tossed a small chunk or two into the "gone-for-good" pile because it wasnt worth the effort to salvage the few ounces of AL from the steel. I ended up with some really nice feedstock from those heads. I spent about 2 hours start to finish.
I've used a lot of different methods of breaking down different types of aluminum scrap and I think this is the winner.
Pete

 

The little I've done, I used a weed burner and heated the Aluminum to hot short and busted it up with a hammer. Got most of the steel out and then melted down the rest and picked out the steel that was left. Fluxed it with some pool chlorine and poured into ingots.

 

Is there a design rule-of-thumb for determining the optimum distance between the crucible and the ID of the furnace?

Discussion - the A60 crucible has an OD of just under 11.5 inches at the top. A 14 inch bore leaves only 1.25” clearance around the top of the crucible. A bit over 2 inches towards the bottom. That seems very tight, but it clears nonetheless. The A20 is just under 9” diameter at the top. Question – would it be bad use the smaller A20 crucible within the larger 14 inch bore furnace? More specifically - what are the negative consequences (if any) from running a smaller crucible within a larger furnace? It seems like the only downside would be wasted energy (heating a large furnace to melt a small pot). On that note, I am highly interested in your low-mass construction method (using Inswool castable with ceramic wool backing, inside a stainless steel sheet-metal liner).

Can you explain what you mean about exceeding wall loading? The cast pockets around the heating elements are exposed to very applicable heat input. The only way I can see to minimize peak temperatures within the pockets is to reduce the energy flux generated by the wire. The OEM recommends 20 to 60 watts per square inch. I calculate that you’ve targeted about 30 watts/in^2. The Inswool castable seems to be holding up. That implies 30 is a good number. Another big variable is the coil spacing. The OEM recommends a stretched coil-spacing of between 2.5 and 4 wire diameters. The wider the coil spacing, the lower the overall peak-temperature should be within the cast wire channels. The obvious limitation being how long you can stretch the wire, and still fit within the bore of the furnace.


Question - is the wall loading concern relating to the hot-face only? My thinking is that if the wire’s heat-flux is kept around 30 watts per square inch, and the coils are properly spaced, I should be able to install 20, 30, or even 40 kW of heating coils into a low-mass furnace, and not thermally overload the Inswool. The PID controller would ultimately limit overall temperature. Thoughts?


On the topic of obtaining aluminum - one of our local wrecking yards has a hydraulic-powered tire remover. It’s basically 4 log splitters pointing inward. They throw a wheel/tire into the center and cycle the rams all at once. The steel rims will come out looking like 4-leaf clovers. Aluminum rims will shatter into small pieces, which they collect & toss into recycling bins. The tires simply fall off. They don’t even bother letting the air out first. It’s a pretty impressive machine. Maybe I should inquire about purchasing the shattered aluminum pieces?

 

I found some design rule-of-thumbs.

Recommended Kanthal A-1 alloy (FeCrAl) heat-flux (for coiled wire simply supported within an open ceramic groove):
30 watts/in^2 for 1500 degree F max oven temp
18 watts/in^2 for 2000 degree F max oven temp
15 watts/in^2 for max 2370 degree max F oven temp
Interesting note: the Kanthal wire guide generically states 20 to 60 watts/in^2. This is misleading. Only very low oven temperature applications would survive 60 watts/in^2 heat flux. Kanthal's guide does not make this clear. Also, the above recommend heat-fluxes are to obtain the longest element life. So what does that equate to? Years of life? Decades of life? Presumably the projected number of thermal cycles before failure. It would be nice to know. Perhaps they intentionally do not speculate. . .

Most common size Kanthal A-1 wire for coiled heating elements in kilns & ovens is 13 to 16 gauge.
Heavier gauges provide longer life at higher temperatures (Kanthal's wire guide specifically advertises higher safe operating temperatures for the larger wire gauges).
Designing for lower heat flux will also increase element life.
Ideal outside coil diameter should be 10-14 times the wire diameter. Note: this is "ideal" and not required. Tighter coils will work, however tighter coils might move around more between cold and hot.
Absolute minimum coil spacing is one wire diameter.
Recommended spacing is 2.5 to 4 wire diameters.
Max acceptable temperature for Kanthal A-1 alloy is 2550 degrees F.
Elements must be completely supported.
They will deform when hot.
Most common support method for coil-elements is grooves cut into furnace walls.
Can also be put inside quarts glass tubes.
Wire becomes brittle once fired.
FeCrAl wire is magnetic, NiCr is not.

Note: When I initially read the term Surface Load on these threads, I thought we were talking about the heat being applied to hot-face refractory coating. Sorry - that confusion was on my end. Surface-load and heat-flux are interchangeable - when sizing heating elements.

 

Your PDF link is no longer good (404 error when attempting to open). I did go to the manufacturer's website. They have dozens and dozens of documents available for download. They did not permit me to download anything without first providing my personal information (not a good business model IMHO). I was also shocked at the number of files available for download - each very limited in subject-scope. You would think that our living within the digital age - with highspeed internet and such - that they could combine the bulk of their downloads into just a few key high-impact files, and openly disseminate them. Nonetheless I did find two of their older wire reference documents available from 3rd party sites (available from vendors trying to sell the wire). As you stated, both differed significantly in content. Neither file identified recommended heat-flux levels vs operating temperature - both documents simply stated a range of 20 to 60 watts/in^2 - which is inherently confusing without further context. Neither document contained thermal break-in procedures for new wires.

Kanthal A-1 wire contains 5.8% aluminum. The recommended thermal break-in procedure is to energize/heat the wire within an oxygen rich environment, specifically to form an aluminum-oxide protective coating on the surface of the wire. Apparently the proper formation of the aluminum oxide coating greatly extends operating life of the wire. It's interesting that a small splash of molten aluminum on the wire will cause it to fail so quickly. I wonder if the failure is a result of thermal shock from wetting the wire, or perhaps negatively impacting the thermal conductivity/emissivity of the wire's surface (causing a hot spot), or if it's simply from leaching in additional aluminum content into the wire (via species migration).

Question - do you ever feed your crucible with additional aluminum while the oven is hot? I'd imagine you'd have to splash it pretty good to reach your elements in a lift-off oven.

I did see a warning from one vendor, stating that cheap Chinese Kanthal wire is being marketed in the US, and to be careful where you buy. Question - how do you know you're getting the quality wire?

 

I found the file you attached. Thank you. And agreed - this is much better reference document (than the first two I downloaded from wire vendors).

Question - why did you not use Kanthal APM in lieu of A1? They claim APM has significant reduction in creep issues, implying it will be less likely to deform & pull out of hot-face spiral-grooves. Apparently APM is still a relatively new alloy - it didn't even exist prior to the 1980s. Older wire reference documents make no mention of APM wire. It seems like a special-order type product. A quick internet search only returned one vendor advertising a 50 pound spools for 80 bucks, in one diameter only. Other vendors wants you to contact them for a customized price quote. It would be nice if AMP wire were as readily available as A1 wire.

 

Click spring made one that I intend to make.