Sciencemadness Discussion Board

Grouted heating elements (rewiring tube furnace)

Fulmen - 27-6-2014 at 03:38

I'm currently rewiring on an old lab furnace, the classic barrel-shaped type. The old heating wire had turned brittle (actually it's crystalline), so it'll have to go.

It's a tube design with a square ceramic tube appr 110x190x300mm with the wire wrapped around the outside and "grouted" in. How do I replace this grouting? Will regular fireplace mortar hold up? Or do I need a special refractory mix?

Will this affect the temperature and life span of the wires? The original wire is 2mm thick and rated at 220V/14A (3,2kW), the new wires are much thinner (0,8mm for the 2kW). It won't see continuous use, but still. Seems very flimsy compared to the original...

Also, the insulation is a light pink powder, never seen this material before. Does anyone have a clue what it can be?

Burner - 27-6-2014 at 05:44

The choice of the appropriate grout will depend on the operating temperature of the unit, which in various designs can range from about 1000C to >1700C. Without knowing the intended operating temperature (and if >1300C, the heating element material) there would be no good way for anyone to offer anything more than general suggestions, which may not be suitable.

Fulmen - 27-6-2014 at 05:56

Good point. I have no idea how high I can go, the unit is labeled 1000°C and that is probably enough for me (it's for heat treating steel). The wire is specified as 0Cr25Al5, which is rated for max 1250°C continuously, but then again Chinese quality control is a bit of a crap shoot.

Edit/Update:
I have found a alumina cement fire clay rated at 1300°C available locally, other than that I have a bit of homemade mullite (ball milled quarts ore and alumina). If needed I'm sure I can get hold of fused MgO, but the fire clay seems like the simplest solution.

[Edited on 27-6-14 by Fulmen]

IrC - 27-6-2014 at 12:01

For some reason it seems few ever actually search the site for what has been previously discussed. Your idea will do poorly, soon cracking apart. The O2 exposure to the element will reduce the lifetime. ITC-100 or 200 at minimum, ITC-213 being the real choice while expensive. I started coating my Kanthal with ITC-213 when replacing my wire for about the 4th time. That was in 2006, and the same (last) element is still working fine in 2014.

http://www.sciencemadness.org/talk/viewthread.php?tid=26899#...

Fulmen - 27-6-2014 at 14:46

Thanks for the pointer, I had seen that thread but didn't think it was relevant. Useful stuff there indeed. Guess I'll have to mix up something myself.

Come to think about it, I do have both bentonite and grog. Trouble is getting stuff like zirkonia, I'm on a very tight budget here.

Burner - 27-6-2014 at 14:48

If you plan to get the IRC coatings this is the cheapest source for them, by far! - http://www.hightemptools.com/itcproducts.html

Fulmen - 27-6-2014 at 15:34

The IRC seems to be a coating, I'm looking for a castable mortar. At least that's the original construction, a solid ceramic tube, coils on the outside cast in a 15-20mm layer of some mortar. Are there any reasons to deviate from this? With the coils on the outside I need a dense, highly conductive material, right? The original insulation is in the form of a fine powder, so unless I change this with say ceramic fibers to create an air layer for the coils I thought I was better off with a solid casting.

Question is, what to use. Isn't mullite fairly conductive?

IrC - 27-6-2014 at 15:47

Yeah that would take a lot of expensive ITC. A thought though, build up a coat of ITC on the elements first (allowing to dry), then use something you can afford for the bulk of the job. Painting the finished work with a coat of ITC when done would improve things as well, reducing the loss from thermal radiation. Also a good idea to study the entire thread I linked before you do anything for ideas.

Fulmen - 27-6-2014 at 15:56

Yeah, I'll read through it once again tomorrow. Spent the evening reading up on refractories but am having trouble finding good data on highly conductive materials.

But I'm having trouble seeing the benefits of a coating if I'm going to cast the entire coil in afterwards. If I manage to get a durable result that should protect the coils from oxygen, right? And a thin coating isn't going to change the conductivity of the setup much anyways. Or am I missing something here?

Fulmen - 29-6-2014 at 01:41

Well, I did a couple of simple experiments yesterday after finding some grog. I wound a 1.5m length of 0,3mm NiCr around a 5mm ceramic rod and covered it with ~5mm of mullite/grog-clay. The forced drying produced a few cracks, but I'm sure that can be solved with more grog and slower drying. I pushed the power as high as 100-150w, causing the wire to burn after a few minutes, but the resulting refractory seemed to have acceptable strength considering it will not see any wear.

Now ideally a grouted in element would be well protected from exposure to air, which should extend life. But at the same time the wire will get hotter, increasing the risk of melting like in my experiment. This makes me unsure about the sensibility of this approach, but I can't see any other reasonable way of doing it. That would entail ditching the ceramic tube and make a new lining from soft bricks, something I was hoping to avoid. besides, they did get this to work for decades with the original design. Granted the wire was much thicker (2mm) and subsequently longer which could keep the temperature down, not sure if my coils can withstand the same. One solution would be to wire the 1kW5 and 2kW elements in series, giving me appr. 900W.

Not sure where to go from here.


Fulmen - 29-6-2014 at 05:13

One thought: The elements I have are coiled, as was the original. There should be room for winding the uncoiled wire onto the tube (spacing appr 12mm), ensuring better heat transfer. I also remembered I have both perlite and vermiculite (is there anything I don't have?), so I could even add a second coating of low-density refractory on top. Any thoughts?

I also redid the experiment with 0,5m of 0,3mm NiCr, covering it with multiple layers of mullite paste to reduce the cracking. It's been running at appr 100W (35V @appr 11ohms) for half an hour now with a weak orange/yellow glow. I'll let it run for a few hours before increasing power further. But this seems to be doable, although I wish I had more information to guide me. I can't seem to find anything relevant on the web, maybe my search-fu isn't strong enough?

Edit/Update:
Here is a build that's pretty much the same as I'm considering:
http://straightrazorplace.com/forge/98230-home-made-furnace-...
And another build here on sciencemadness:
http://www.sciencemadness.org/talk/viewthread.php?tid=14267&...

So it seems I'm on the right track. Question is what to use to cover the elements. I'm leaning towards a thin layer of mullite with an outer layer of high-temp hydraulic mortar for rigidity.


[Edited on 29-6-14 by Fulmen]

IrC - 29-6-2014 at 11:21

I would go for building it with materials you have and finding out through experience if it lasts a reasonable time. On my kiln the elements are exposed to the atmosphere inside the kiln and over 2 years I had to completely rebuild it every 6 months (everything I was doing was cone 10 at minimum). The 4th rebuild was in 2006, where I used about $150 worth of ITC. 100, 200, 213 in various areas. The 213 used for total protection from oxygen around the Kanthal, and as an overall 'reflector' for the kiln walls. My troubles were not so much how hot the wire, rather how well protected from O2 it was. The inside and outside thick coat of the entire brick structure with 213 made an enormous difference in cycling. I think this may have been of the greatest importance. In the prior use the kiln was heating nearly all the time. After the 213 work I could see it cycling more often. Clearly radiation loss was a big problem which forced the elements to be powered nearly 100 percent of the time in all prior runs. If anything has saved the element I have to think it is the fact there is more 'off time' during a run. The 213 coating the elements did not reduce heating, when up to temp I believe the 213 is radiating heat from the wire very well. While being no expert I did conclude through experience O2 was a bigger problem for hot Kanthal than was the temperature of the wire. Possibly I'm wording this poorly but what I mean is being at a white heat is less of a problem for the wire than is being this hot and also reacting with oxygen at the same time. I see your point that being well grouted (even with cheaper materials) may give your wire enough cover from the air. In any case my conclusions were how hot the wire operated was less of a problem than how much Oxygen it was exposed to while at a white heat, and the better the reflectivity of the walls the less time the elements were on. Both parameters work together to give me a hotter kiln which lives long and is cheaper to operate.

"So it seems I'm on the right track. Question is what to use to cover the elements. I'm leaning towards a thin layer of mullite with an outer layer of high-temp hydraulic mortar for rigidity."

This I cannot answer having no experience with these materials. But I can give insight that it is important there be no cracking over time which would allow O2 to reach the hot element. At minimum inspect the kiln every so often after it cools down for any sign of this.

Fulmen - 29-6-2014 at 13:08

Thank you for the feedback. I must admit I haven't understood the function of these coatings fully, especially how such a coating could increase efficiency without increasing the insulation value. But as for O2-protection your experiences seems to fit with mine. Lab muffle ovens like the one I have can run continuously for decades (mine certainly has), although not at cone 10 (1300C).
Having seen other builds with grouted elements I am more confident that this will work, sadly the design doesn't allow for easy inspection of the elements. It's one of these:
http://upload.wikimedia.org/wikipedia/commons/3/3d/Muffle_fu...
In order to get access to the elements the front must be removed and all the insulation dug out, so it's a bit of a chore and adds the risk of damaging either the leads or the grouting. That's why I want to get this right the first time.

IrC - 29-6-2014 at 13:44

I see your point inspection is not really an option. A thought would be get a pint of ITC-213. Install the element wire and then coat it using a brush. After it dries well finish doing the grouting and building the kiln. This way if cracks develop in your grout there is still a permanent barrier against O2 on the wire itself. The cost would be low as you will not even need the whole pint of 213 and it will stay coating the wire no matter how hot you run it. The wire itself would melt before the coating deteriorated and you are never going to run that many amperes through the wire so it would seem your job would last years.

Fulmen - 29-6-2014 at 14:06

I will do a search for that or similar products. I also discovered I had a can of 2000F exhaust paint, while it can't withstand direct contact with the elements I could possibly coat the mullite with it after one firing in the open. it produces an enamel-like coating that should work as a barrier. Doesn't seem to be any OTC paints that are any better, but I'll also put out some feelers for industrial products that could work.

Fulmen - 30-6-2014 at 00:15

Jeeez, I must be blind. Top sticky was a 6-page thread about tube furnaces. Completely missed that one, have some reading to do tonight. Already found a reference to Brauer where waterglass/talc-coatings are recommended for temperatures up to 1000C and MgO/Al2O3 for temperatures above that. It also warns against silicic acid attacking the elements at elevated temperatures, so the mullite might not be such a bright idea after all.

Fulmen - 1-7-2014 at 02:07

This wasn't as simple as I thought, been reading feverishly for a couple of days now. Sadly, the more I learn the more uncertain I'm getting :-)

Might as well start with the things I have decided on:
Elements: I'll be using the 1500W and 2000W elements, wound in parallel around the tube (appr 15-20 turns each) with a "center tapping" configuration. This allows both individual, serial and parallel configuration. I will try serial first, giving appr 900W. If this doesn't provide enough power I'll probably try parallel with a dimmer or an inductor to regulate the power down, I figure this will extend element life. Although oxygen exposure seems to be the biggest factor when it comes to life span keeping the temperature down can't hurt either.

As for coating I will have to do a little more digging. The talc/waterglass seems like an interesting choice, although I'm not completely comfortable with using sodium silicate directly on the elements. The MgO/AlO-coating seems safer except I'm worried about the structural strength of this mixture without any bonding agents. Or am I over-complicating things here? If Brauer recommends the talc/waterglass for up to 1000C the higher temperature of the elements must have been taken into consideration, right? 1000C won't permit heat treating of special steels like HSS, but I think I can live with that. The challenges seems to increase exponentially with temperature...

As for the outer refractory I'll try a couple experiments with phosphoric acid as a binder for the mullite, if not the 1300C refractory mortar seems like the best solution. As long as I coat the elements with a decent layer of talc/WG the mortar ought to hold up.

Simply buying something like ITC213 seems more and more sensible, sadly the shipping kills the budget (80$ total for 1/4 pint from higtemptools).

[Edited on 1-7-14 by Fulmen]

Fulmen - 3-7-2014 at 06:15

The talc/waterglass coating seems promising. I mixed up a small batch and applied it to my test-resistor (0,5m of 0,3mm NiCr, 10ohm), and it produced a semi-molten, glassy coating. But to be fair, the homemade water glass was probably very high in free sodium hydroxide which could lower the melting point. But even in this soft state it holds up well, the resistor has already been running for an hour @ 100-150W (glowing orange in daylight) without it burning out.

I will continue the testing for a while, if it holds up I think I have found my choice of coating.

Burner - 5-7-2014 at 11:35

Just a minor caution to consider - "glowing orange" is ~900C (~1700F) which is significantly cooler that your expected wire temperature of ~1250C (~2300F). Reactions, diffusion, etc. increase dramatically as temperature increases, so performance of a coating or system @900C is not necessarily a good predictor of those materials @1250C.


Fulmen - 5-7-2014 at 15:41

The entire coating was glowing, obstructing direct view of the wires. And considering the heat loss the wires must have been significantly hotter. I'm going to trust a reputable source like Brauer this time. I redid the test with a new batch of water glass, and this time it didn't melt nearly as much, so I guess it's time to shop for supplies.

WGTR - 6-7-2014 at 09:16

Your mileage may vary, and this may or may not be good advice... but I would think that whatever coating is used, it can't remain molten at normal operating temperatures. This is because your heating elements rely on a tough oxide layer to protect them from further oxidation.

If a molten flux is allowed to remain on the wire surface, oxygen diffuses in to the heating elements, the protective oxide layer is dissolved, and then element lifetime is decreased.

Perhaps a volatile oxide like MoO3 will help here. It melts at 795C, and sublimes at elevated temperatures (1100-1200C). This means that after it does its job, it volatilizes away.

aga - 6-7-2014 at 14:05

Would it not be possible to make a furnace where the heating elements are in a vacuum ?

No O2, so better/cheaper ?

Oscilllator - 8-7-2014 at 01:08

aga if the heating elements were in a vacuum, then I imagine they would heat up and melt fairly quickly, because of course a vacuum is a perfect insulator so the only way the wires could lose heat is through radiation.
Aside from that, there would be enormous technical difficulties associated with creating and maintaining the vacuum with such high temperatures

Fulmen - 8-7-2014 at 01:40

WTGR: The melting was most likely caused by the homemade water glass. I ended up with a mix of 50/50 talc and fine grog bound with a commercial concrete sealer, and this holds up just fine. It fuses upon heating, but doesn't melt like the first experiments. I'll have it done in a couple of days, then we'll see how it pans out.
Question is whether I should add more refractory on top of this or not...

Fulmen - 8-8-2014 at 05:55

Woohooo! First full test run, and it got to 800°C in 2,5 hours at 900W. I'm sure I can push it to 1000C, but I will try to minimize that as much as possible.

Also did my first case hardening of low-alloy steel and it came out nice and hard. This is going to add some possibilities for the future.

Fulmen - 10-8-2014 at 10:57

Second run today, testing out case hardening and testing some mystery steels for hardenability. Connected the original thermocouple, best fit seems to be S-type although it does differ somewhat from the readings with the K-element that came with the thermostat. Any clever tricks for identifying a TC?
Also, how much can a 30cm copper extension wire screw up the measurement? The junction is pretty much at room temp.