Wire Melter Melt Down....

[Garner Britt]

Ok, here's what happened

-Built a 75 lb wire melter
-Bought a 75 lb pot (thanks Pete)
-Got some SP cullet (thanks Glenn)
-SLOW ramp up to 2000, and charge warm cullet
-After getting it about 3/4 full, decide to go inside and drink a beer.
-Fall asleep in chair in front of t.v.
-Wake up some hours latter and go out to put in the last charge before going to bed.
-PANIC, controller is showing a burnt out thermocouple and furnace seems really hot.
-Replace thermocouple with another brand new
-Controller shows temp of 1600f and falling
-Looks like I've lost 2 elements.
-Mess with wiring, check relays, etc. still falling
-Grab pipes and punties and start trying to drain the pot.
-Realize there is no way the furnace is only 1600, I can gather.
-Temp continues to fall I can gather all the way down to 1300 (on the controller)
-Get the vast majority of glass out and let er come down slow.

About 30 hours latter, I can pull the crown and the pot and check the elements.

here's the list of symtoms

1. Broken Alumina thermowell tube
2. Melted ceramic and metal from the first themocouple.
3.one element just broke, like they do
4. 2 of the elements have really burnt up, sections 1-2" totally gone, replaced by molten element fused to the brick.

What do you think happened? lose a thermocouple, over heat (this thing's got (had) 8800 watts) burn elements up, second thermocouple reads incorrect do to shorting out on goo inside thermowell?

I do know that I won't be trying to get away with the cheap type K thermocouples anymore. Any advice on finding thermocouples that will hold up but don't cost $200.

I had planned on switching all 4 elements so I could run it on 2 or 3 and save an element as a back up. This might have prevented an over heat (if that's what happened)

I guess over temp protection is worth will on a wire melter after all.

Any suggestions will dealing with the molten elements in the groves. I haven't tried anything yet but I'm sure its going to be a pain.



garner

[Brian Gingras]

more details..

what ype of T/C tube were you using?

what kind of bricks did you make this with? low iron?

what kind of load did you have on each element (watts per Sq inch)?

[Garner Britt]

TC = cheap ceramic bead bare wire supposed to be good to 2350 F. Replacement TC for cone 10 ceramics kilns. Inside an alumina thermowell tube. I see the errors in my ways already on this choice.

Bricks are Harbison Walker IFB 2600 LI (low iron) .8% iron oxide. though there is a distinct red glazing on the bricks after this mis-hap.

The elements are off the shelf from Duralite. 2200 watts /9.2 amps at 240 volt. Kanthal A-1. Not sure of the watt loading. They are pretty thin gauge wire probably 16 or 18 gauge. Started at 38 inches stretched to 120". None of the failure points were at the element pins. I do know some beefier elements would be better but thought I'd get by with these for a while, and they're cheap.

garner

[Steve Stadelman]

Well, If I was you I would contact Scott Dunahee and change over to 11ga. Unfortunately noble metal t.c.s are just freakin' expensive. However, my friend Kenny Simpson makes 8ga "K" type for about $25.00 that will at least last a couple months in an alumina sheath.

[Garner Britt]

Sent email to Scott,

started some research on designing elements. Tell me more about watt loading and how it's calculated. I had assumed that Duralite would have designed that into these elements.

Does my guess as to what happened make any sense?

lost tc signal, experienced what Fuji calls "unwanted continuing heating" (over heat) then basically found the melting point of Kanthal A-1?

garner

[Brian Gingras]

I think what we have here is a learning experience. First off, those elements are not bad for an annealer, but way to light weight for furnace works. The thermocouple should be sheathed in an alumina sheath.

If you hit the melting point for the elements then I suspect that you also went way beyond the rating of the bricks, castings, etc etc.

watt loading is calculated as watts per square inch of element. It's all in the archives, I would be reading all of it if I were you. Pay close attention to the watt loading recommendations.

[Garner Britt]

The tc was in an alumina sheath (thermowell) though it was only 14 ga wire, I'll try 8 ga next or some noble metal tc if I can find a deal

I've used these exact elements in 2 cone 10+ (2300f) ceramics kilns which have held up fine. They don't stay there long but still. Kanthal A-1 is rated to 2350 according to duralite.

Obviously I need some beefier elements but I'd think I'd have gotten a few cycles out of these while I sourced bigger elements. I still think the weak link was the cheap-o tc

I'm not too worried about my refractory, it's a wire melter after all and designed to be rebuilt as needed. Kastolite 30 and 2600 degree fire brick should be fine after a breif trip to 2350 or so. (except for the melted elements deposits in the grooves)

In my humble opinion, the archives combined with the search function of the board makes for some great information that is nearly impossible to find.

garner

[Richard Huntrods]

First, I think your assessment is correct - the TC melted, the controller just kept applying power until the elements melted.

I agree that 11Ga elements is about as thin as I'd go.

BUT - my real advice here is to bite the bullet and purchase a good type "R" thermocouple in an alumina sheath.

Type K is just not rated for glass melting heats, and if (when) they fail, you have seen the results.

I'm guessing (based on my own first learning experience) that you were well above 2400F (been there, no melting like you describe).

My thermocouple was bought from WICA - Alltemp (canada) and is a:

CPTAA-1X11W-2C-E-26-1-R-12-10.5 (type R thermocouple + head)

In 2003 (Sept) it cost me $400 CDN. It would probably be less now for me.

Cheers,

-Richard

[Drew Fritts]

Although they appear to be an expensive initial investment, the $210(+/-) I spent on each R-type thermocouple with alumina sheath and weather head from Denver Glass has been well worth it in the long run. I've been using the same two for 8 years now in my wire melters.

Drew <+><

[Victor Chiarizia]

try Kulas systems in somers ct. use to get all my stc's from him. nice guy with a small 2 man operation. i have some old pt thermos and will think about selling them. vic

[Dave Bross]

Best way is the type R thermocouples... but if you have to cheap out, McMaster Carr (mcmaster.com) sells an 8 gauge type k. I get about 3 months out of each one in a small melter. You can tell they're going out when the numbers on the controller start to wander at random. Easier to just replace it at 3 mos.. You can get the tube from Duralite or most pottery supplies.
I've used the mullite tubes with the type K and been OK.

If the bricks aren't totally crumbly you can get what junk you can out the element grooves then get some RFC 17 furnace coating from wesbond (wesbond.com) and paint everything inside the furnace with it. You can mix it up thick and cover all that leftover trash in the element grooves with it.


Watt loading is the watts (volts multiplied by amps) that the furnace draws divided by the surface area of the wire.

The wire can be thought of as a very long cylinder so you can use the formula for the surface area of a cylinder to find the surface area of your wire. The formula is:

2 * pi * radius * length

* means multiply
pi = 3.1416

radius is half the diameter

I've had good results with a watt loading of 8 watts per sq. in. or less, but it takes a lot of wire to get that. The thicker the gauge the more wire you need and it's tough to fit all that wire in the furnace and still have a reasonably small chamber. You also want 2000watts per cubic foot so the less chamber size the less watts you need.

Ed Skeels wrote a great little element calculator program, you can download it from his website, Keith Orr blowpipes. It's really handy to tinker with element sizes. You'll need the Kanthal chart of A1wire diameters and resistances to use it and I don't know where you can find that offhand.

I agree with the 11 gauge suggestion, but I personally do a compromise using 13 gauge wire, 8 watts per sq. in., and only expect 4-6 months element life. Thinner wire has more resistance so less of it required. I have a one cubic foot chamber running about 2400 watts and it melts batch like a charm, but no room to put the extra length wire something like 11 gauge would require.. This element life coincides nicely with having to swap out the thinner wall EC pot that I use at about that 4-6 month point, so in my unique situation, it works. This (obviously) is also a much smaller melter than yours.

Running elements in parallel drops the watt loading but it also roughly halves the resistance so you need twice the wire...and a place to put it. I think that's why a lot of folks melting in kilns are having good success with thin gauge elements, but many elements in parallel.

[Brice Turnbull]

I can't quite solve your problem from here, but will tell you what I can, so maybe you can solve it there.

16 or 18 gauge A-1 is too thin to last very long, but should still last (or at least it used to last for me) a good 5 or 8 months of continuous use never to exceed 2280 if you ramp up slowly - IF properly matched to your power supply.

You were right that the skinny wires should have gotten you started, I don't think that caused the failure; unless duralite made a boo-boo and sent you the wrong resistance worth of wire. Duralite was good to me, but every human is still capable of a whoopsy. Also (without having been with you when you ordered them I can't know) it may be possible that you gave them incomplete information when you ordered the elements.

Also, type 'K' tc's (as a couple of people have mentioned) should last 3 months or so, so I don't think that is your immediate culprit either.

Does your controller allow you to chose the type of thermocouple (K,R,S) you're using? Perhaps it is set to the wrong type?

It sounds to me like your thermocouple and controller were not communicating for some reason.

Just a hunch.

The suggestions for type R and 11 or 10 gauge Kanthal wire are good, however you are correct that it should be functional (knowing you'll have to replace tc's and elements a couple times a year if you run constantly).

On the sunny side of the street, if you find the problem and solve it and it does not ultimately involve the elements or tc directly, while I don't envy your jittery emotional future over the next few melt cycles, at least you lost the cheapest possible elements in the easiest to repair furnace you could have :)

Cheers, Brice

[Garner Britt]

Thanks everyone for the advice so far, I was obviously trying to get off cheap and, well, didn't.

Heres my current plan of attack.
1 Get a better tc. I will try a $15 mcmaster 8ga K in an alumina thermowell.
2 replace the 3 bad elements with off the shelf alternatives from duralite.
3 run only 2 at a time to avoid over-temp.
4 piece together an overtemp shut-off with separate tc, controller and relay(s) (with parts I have on hand)

In the near future, I will buy or wind some 11,12,13 gauge elements. buy a real tc, when I can find a good deal. and spend less time worrying.

I spent enough time building this thing, I'd just be satisfied to get it hot and blow glass and actually do it right (=$$$$) later......I know, impulsive and fool hardy but......

thanksOriginally posted by Dave Bross
You'll need the Kanthal chart of A1wire diameters and resistances to use it and I don't know where you can find that offhand.

you can find that here:
http://www.resistancewire.com/mainpage.php?page=gaugechart

click on the "Click Here to Read the Datasheet for This Alloy, KA1"

thanks again and keep it coming.

garner

[Garner Britt]

Also,

The Fuji does allow tc calibration selection, It is indeed set for type K.

Do any of you think that excess mortar in the grooves could cause problems. It contains 1.4% iron oxide. I mortared this thing together (with Sairset(I know, it's crap) and any excess mortar in the grooves was sprayed with a mister and brushed down flat and thin....just wondering.

garner

[Brice Turnbull]

Oh, I thought you'd had Duralite wind custom elements for you. That's what I did for several years until a friend offered to wind some for me.

Call Duralite, tell them the specifics (volume of space to heat, how hot you want to get, and wire gauge you want them to use - maybe some other things they will ask). Be ready to order 3 or 4 elements custom wound by them for you. They should connect you to a tech who will figure the element specs with you. Then you pay them to make you a set or two of elements (it's good to have a spare set).

Cheers, Brice

[Glenn Randle]

Originally posted by Garner Britt
Ok, here's what happened
-After getting it about 3/4 full, decide to go inside and drink a beer.
-Fall asleep in chair in front of t.v.
-
garner



That was an expen$ive BEER!!

I bet it was the same brand that caused me to propose!! :headbang:


I wish I could sue the Beer Companies for all the stupid stuff it's caused me to do over the years. But then I'd probably have to prove that I don't do stupid things when I'm not drinking....., that's the catch!:feathers:

[Garner Britt]

Originally posted by Glenn Randle

I bet it was the same brand that caused me to propose!! :headbang:


Atleast she lets you keep an oil burnin' bucket o' glass.....

garner

[Marcel Braun]

Hi Garner!

If your K type TC shorts to the outside of the box it will start to read a temperature in the area between the outside and inside of the furnace, or at least a falsely low reading.

Is it possible that one of your TC leads shorted to the outside?

I lost a weeks worth of boro hotshop work to this while reannealing a set of components after coldworking and flame polishing.

-Marcel

[Garner Britt]

hey Marcel!

The first TC burn up, leaving metallic goo in the alumina sheath. I think the second was shorting out on that goo and giving me bad readings (in this case colder than actual)

I've been thinking that although this was a pain in the ass and a bit expensive. It's actually really fortunate that the elements gave out and melted. I wonder how hot I would have gotten with 11ga elements and a failed tc....

this wimpy elements are like fuses.....

garner

[Dave Bross]

I don't know what you're switching the elements on and off with, but if it's an ssr they do tend to fail by staying "on" all the time.

I would suggest not mortaring the bricks next time so you can replace individuals when needed later. The RFC 17 stuff from wesbond will seal off most anything. My best guess is that it's a colloidal alumina with zircon, the zircon being your basic bad-ass mineral...hard likie diamond and impermeable to most things.

Wire melters are high maintenance, so expect more fun and games, but it's a very inexpensive way to get hot glass despite all the aggravation. Sort of like driving an old "beater" of a car.

[Garner Britt]

Originally posted by Dave Bross
I don't know what you're switching the elements on and off with, but if it's an ssr they do tend to fail by staying "on" all the time.


Good point, but they are fine, 50 amp units for a 34 amp load, heavily heatsinked, fan cooled and located well away from any heat source....

I still blame the tc and I think overtemp is a really good idea on wire melters. It's true they are sort of self limiting but in a destructive sort of way

garner

[Kraig Richard]

make sure your wire spring is really well streached out, arcing betwwen coils can F up an ellement. be sure to double up and twist the wire at the ends of the element. ass that keeps them from heating up where doubled. Omeaga has really cheap thermalcouples and tubes.