Type of Thermocouple for Molten Aluminium

Why go for contact measurement when there are other ways (rhetorical question)? There are laser-based, non-contact thermometers available. One shines the laser dot at the surface, and one receives a read-out on the display on the instrument.

My concern would be durability. This is a foundry with hot metal being splashed about. I've done a few similar jobs and my experience has been that if it's not tough it doesn't survive.

When it comes to unattended temperature control I'd be very nervous about straying far from my true and tested method of two thermocouples. Each with it's own display, one driving control, the other driving over temperature shut down. Problem is what to house the thermocouples in. The previous, similar pot melters that I've done have been for tin and lead and apparently the iron contamination issue isn't a concern.

I'd love to go for non contact but I can just imagine it looking at the floor because some clown nudged it, accidently or otherwise, with a forklift. Meanwhile the metal and it's expensive crucible are being cooked. Thermocouples in a steel (not in this instance) tube bolted into place are cheap, effective and almost idiot proof!

Yup, clowns are known to do that. That way to avoid that is to house the thing in such a way so that it won't get bumped or hit.

We installed our infrared thermometers by cutting in a hole in the wall and putting a glass window in front of it. When the quality assurance guys objected to having potential glass contamination of our product, we removed the glass and put the device inside a tube and injected cool air into it. Since it was in a recess in the wall, it stayed out of the way and the cooling air kept it from overheating. 'Been fine for several years now.

Hmmm, it'd be a very neat solution if I could make it 100% safe. My concern stems from my intention to nearly double the burner output as part of this modification. Currently I've got it throttled back so that if the operator is not up to speed not too much damage happens. Now I'm handing control over to a PID loop I feel I can open it up a bit but I've got to be sure that in the event of a component failure the over heat will still kill it.

What sort of process are you using your IR sensor on?

Infrared thermometers are good for this. They're non-contact and fairly accurate as long as you adjust them for the correct emissivity of molten aluminum.

You can buy those types that provide a 4-20mA output. There are other types available that provide the same mV output of thermocouples so you won't need to change your equipment. Just google "infrared thermometers". Raytek and Omega offer industrial types.

If you must have a contact thermocouple use a type k

It consist on one platinum wire and one mixed platinum/rodium wire.

Put "type k thermocouple" into your search engine of choice and read the results. rbeadle@pcc.edu

Any suggestions as to a material for the housing?

Type R and S are platinum/platinum-rhodium - Type K is alumel/Chromel.

Platinum + 10% Rhodium/Platinum is Type S, not Type K. Type S is the highest temperature thermocouple available (up to 1750 deg. C according to my table). That's not the problem though. The problem is that the sheath/tube/thermowell cannot be left immersed in the molten metal because it will eventually melt and contaminate his product.

If stainless steel thermowells are not acceptable, other materials are available.

I Googled Titanium thermowell and found:http://www.titanindia.com/thermowells.html

I know Omega has zirconia wells too.

I also know of a company that coats metal parts for chemical resistance using silica, but don't have link on hand, but this may also work for you.

Try this link

http://www.mikroninfrared.com/literature/Aluminumbrochure.pdf

Thanks for that. I had a look. I notice that they recommend that for molten aluminium it needs stirring to get an accurate reading due to the oxide film.

This is the sort of concern I have with non contact measurement. Vulcan in the posts above recommends IR measurement but I'm not entirely convinced. These melting pots are filled with all sorts of different forms of metal, from ingots to shavings, as it melts the surface is constantly changing until it's a totally melted shiny surface. I just wonder whether the changing emissivity is going to totally upset things. The attraction of a T.couple in a well down near the bottom of the pot is that it really doesn't matter what's happening up top, such as some fool jamming a big ingot in, that isn't in contact with the molten material below.

I'm certainly looking hard at all the different suggestions though. Possibly a combination may end up as the ideal.

Do you do your melting in batches or is it continuous? I'm not familiar with aluminum production or recycling. It sounds like you're doing it in batches. You put your ingots or shavings in the pot, melt it and then pour it out. The difference in state (solid and liquid) will pose problems as you say.

Why not give those guys a call? They claim to have been supplying the aluminum industry for years so they might have a solution already.

I can't tell you what exactly we used the IR measurement for. The output of the process is also molten but that's not what the thermometer was measuring.

I'm not fully across the process myself as I'm just the bloke who's called in. It's a batch process which produces alloy for bearings. They heave the raw material in and ladle out the molten metal into moulds. It takes about two hours to melt a batch whilst an operator stands around scratching himself. The object of the upgrade is to automate things so that it can be loaded and left to it's own devices with an operator returning to ladle out and reload. When they are not working 24 hours it can also be loaded in the evening with a time clock firing up in the wee hours so the operator arrives to molten material.

You can understand that I'm nervous about anything less than 100% reliability in the temperature monitoring area. Burners that light when no-one is there need bullet proof control systems!

At Hi-Temperatures Contact-less Sensors [Colorometer] is used.

If you have ever visited a Blacksmith's shop & seen him Hardening/Tempering a steel tool, you will notice when the piece is red/white hot he dips it in water just for a second or ?, & observes the color of piece to change & dips it again when he notce a certain colour. This all is the conventioal COLOROMETERY, used in modern Tech now.

We too have had this problem where type-k thermocouples are rapidly degraded in molten aluminium. We have solved the problem by cementing an outer alumina ceramic single bore capillary tube over the thermocouple and cemented it in place.

The thermocouple now has exteneded life and the response is not compromised.

Mike Stackpool (Applications Manger,Ionotec Ltd) mike@ionotec.com

We use the ceramic thermowell for our furnace T/C which is a type K or type R depending on the range measured. The problem with ceramic is its brittleness. Prone to breaking if abused. Tantalum perhaps could be used if need be for the sheath. It melts at @6000 def F.

The cold leg of your pyrometer can be standard 1/2" sch 40 pipe. The hot leg (the portion submerged in the aluminum) must be a material that can survive for a long period of time. Check with your thermocouple supplier. They should have an exotic metal like Haynes 556 which can survive in aluminum. If they do not have a metal alloy which they can supply for this application, then they will probably have a silicon dioxide or ceramic type of material. The primary disadvantage of the non-metallic option is the susceptibility to mechanical shock and thermal shock. The mechanical shock will come when the operators smack it while drossing the kettle. Thermal shock means you will have to heat it up slowly until it is approx. the same temp as the molten aluminum before you place it in the kettle.

I would stay with type K. It will last longer than the protection tube anyway. Even with an exotic alloy protection tube, you should not expect a life span of more than one year.

possible sources Watlow or Pyromation. You might do a search for Haynes. They make several alloys for numerous applications.

How about a long molybdenum cylinder. Bore a hole down the center, not all the way through, to accommodate the thermocouple. To make a good thermal contact between molybdenum and thermocouple deform the cylinder somewhat with a few blows from a hammer. The molybdenum is now between the aluminum and the thermocouple. Take care that @ 1000 deg C the molybdenum will get brittle after some hours and has a reduced yield-stress (but still enough for comfort). Information on molybdenum is available at manufacturer Plansee in europe.

On this application I had the same problem you are emntioning, the best results for using thermocouples was to use a Temp_Pro ceramic thermowell; however this is expensive and fragil, extreme care must be taken not to brake the well when it is hot.

A non contact radiation thermometer could be a better application. I suggest you contact your local Ircon distributor. You can measure the inside tip of the ceramic well with an IR thermometer. You do not have to remove the well, or install the temp sensor in a retractable cover or lift it up at the end of the batch, with out dissambling the sensor.

Many years ago, I used similar type thermocouples in steel sheaths to measure coke oven wall temperatures in operating coke ovens.The use of a type K thermocouple is likely the right TC for that range. Later, the TC protection was changed to a ceramic tube (similar to what you mentioned). Of course in our case there was much vibration and shifting of walls which can snap the tubes, but use of a ceramic tube may work if it is installed where it wont be purposely put in harms way.

The real issue with a ceramic TC protective sleeve is response time because of the reduced thermal conductance of the material compared to steels. BTW, there are several high temperature steels that can be used as protective sheaths, I am not a metallurgist, but I believe you might consider type 446 for that application.

I've been looking long and hard at the non-contact option since first posting and I've pretty much decided against it. It's been an interesting investigation though, learnt much about IR. I use the handheld units quite regularly and now know why the results vary.

I think it's coming back to a metallic thermowell. I don't believe ceramic will survive. My question is now what sort of metal. I'm having trouble resolving whether or not stainless is going to affect the product. If it is I'll be looking at some of the more exotic options mentioned above.

Answer: The device you seek is made by Pyromation, it is called Cerite.

Cerite? That's a new one on me. I presume that it's a material for thermowells. Thanks, I'll investigate further. Currently I'm looking at 601 Inconel and Sialon, both of which I can obtain locally. Has anyone any experience with these?

Sialon is the answer to your ORIGINAL question. Do not use Inconel, do not use 446, and Cerite is cheap and crumbles. Use a type K, 8 gage with ceramic beads, all inside a true Sialon tube. This will provide you year+ if you take proper care of the tube. Watch out for drifting of the type K - always use at ~same temperature range, not lower.

Many thanks. My customer has now ordered a sialon sheath for this application. I'll post a follow up when it's all operational and we see how it survives. Might be some months away yet as there's a lot of approvals and paperwork to go yet.

we use a silica sheath and k-type thermocouple;for greater robustness you can encase an inconel sheathed thermocouple.The key to non-reaction with the melt is to dip the silica in white emulsion paint-the Titanium Dioxide then forms a non-reactive layer and we have them immersed for months with no degradation