PDA

View Full Version : Furnace Question:

Teenage_Machinist
12-28-2009, 02:28 AM
I have made a small electric furnace with nichrome wire. It's in a gallon paint can, Kastolite refractory with kanthal wire in grooves. It uses 700 watts and has a chamber 3 inches dia, 6 high.

It seems not to be getting very hot- small metal items barely glow. I have not been able to melt a pyrometric cone # 06 (1800 F),

Do you think this is underpowered? Underinsulated?

Black_Moons
12-28-2009, 02:34 AM
Its both. given an insulated chamber, max temp is the tempature where input wattage reachs disipation through the insulation. 700W sounds pertty low reguardless.

dp
12-28-2009, 02:53 AM
What is the temperature of the heating element? If it is not greater than 1800º then there is no possibility to heat the chamber to a higher temperature.

Does the outside of the container get overly hot? If so it is shedding heat and that is a problem. Consider putting it inside another larger can that has R19 insulation around it on all sides.

700 watts is an energy level but not a heat level - recall the entropy issue. Heat moves from a hot thing to a cooler thing. The attainable heat in a low loss system then is the temperature of the hot thing. It is possible to design a machine that burns 700 watts and barely rises above room temperature, for example.

A glaring example of watts vs temperature (pardon the pun) is a heliostat mirror system that focuses heat gathered from a large area on to a small area. When the heat of 100 square meters is allowed to heat a parking lot it gets pretty warm on your bare feet, but when those 100 square meters of energy are focused on 1 square meter the temperature soars. It will melt your Reeboks.

Your system is shedding all those 700 watts at a temperature that is lower than desired. It will require excellent insulation to prevent the energy from escaping before the temperature reaches the desired point.

In theory, with perfect insulation, energy input just 1 watt in excess of energy loss will eventually create enough heat to melt it's container. That is obviously difficult to achieve. Much easier is to user far more energy and to be intentionally wasteful.

winchman
12-28-2009, 07:43 AM
I don't think adding insulation around the outside of the furnace is a good idea. Sure, it'll raise the inside temperature, but it'll also raise the temperature of the outside of the furnace. It may get hot enough to melt the insulation on the wiring, or burn the outside of the furnace itself.

Is this a commercial or a homemade furnace? Got a picture?

Roger

CountZero
12-28-2009, 08:07 AM
Have you made it yourself? If so I would start with at least doubling the power, or triple if you have the means to power it. Some extra insulation could be a good idea too.

Duffy
12-28-2009, 10:58 AM
I have a small commercial unit that draws 1500 watts max. and heats to over 1700^F in about 10 minutes. The chamber volume is approx. 144 cu.in. for a power of about 10watts/cu.in. Yours, on the other hand, draws 700 watts and has a volume of 42 cu.in. for a power of nearly 17 watts/cu.in. I conclude that you are losing too much energy to radiation, and thus you need more insulation. Duffy

snowman
12-28-2009, 11:08 AM
Your power is fine, you need better insulation. Specifically, you need something that reflects the heat, like insulating fire brick or kaowool. You could honestly do what you are trying to do with 2" of kaowoll tied to the inside of the paintcan with steel ties. Later in the year I can send you some kaowool, I just don't have time to dig it out right now.

Pyrometric cones are not a measurement of temperature, but are a measurement of time and temperature. When I fire ceramics, I can have a load reach 2000 without bending a 06 cone, simply because I fired it very very quickly. If I am firing a load and expect to use the kiln sitter (monitors the cone bending rate), I fire very slowly. It will take me 8 hours to get to 1800 when the cone deforms enough to trip the trigger.

You need a cheap thermocouple/meter configuration. Can be picked up on Ebay cheap.

JoeFin
12-28-2009, 11:19 AM
Teenaged Machinist

I know where you can get all the fire brick you want a "Day Trip" drive away from you

mechanicalmagic
12-28-2009, 12:40 PM
Do you think this is underpowered? Underinsulated?
Mainly underpowered, raise it up to 1500 watts, and lower your heat rise expectations. My 1500 watt oven takes 1 hour to get to 1550 f. It has 2.5" insulation all around.

wierdscience
12-28-2009, 12:50 PM
Your gonna need to up the wattage.Our small heat treat oven at work is rated at 5600 watts @ 220vac.

What does your nichrome mfg recomend?Like said above if your element never reaches 1800 the furnace never will either.

whitis
12-28-2009, 12:55 PM
I don't do casting so I will focus on the physics of heat transfer.

1) Cover the top with more refractory material to keep the heat in as this may be your biggest loss. The sky, absent clouds, is near absolute zero. Heat loss due to radiation is proportional to the fourth power of the temperature difference (and proportional to the effective surface area radiating). At 800C, radiation to the sky would be several times higher than to a room temperature roof (not that it would stay room temperature for long!), everything else being equal. At 400C, the sky temperature increases losses tenfold vs room temperature. Outer space is awfully big, don't try to heat it. Plus covering reduces convection losses.

2) I assume the bottom is insulated, if not it should be.

3) Add insulation. Given that you are starting with such a small can you probably don't have much. remove the outer can and replace with a bigger one. Though you can probably add as much insulation as you have inside the can to the outside of the can without melting the can (paint is another matter and the can may weaken) as the can will be halfway between the inside temperature and the outside temperature. Besides, your nichrome will melt at around the same temperature steel does so you aren't going to be running the inside hot enough to melt the can even if you add substantially more insulation. Thermal expansion of the steel might put cracking pressure on the refractory material surrounding it, though.

4) Increase wattage by making the wire shorter. Remove turns primarily near the top, not the bottom - heat rises. If you use half the wire, you may get about 1500W. However, you may want to reduce the amount of wire gradually as you want a temperature high enough to melt your work but not high enough to melt the nichrome. You might want to add insulation first as it is harder to put wire back after you have cut it of, if you do not have an external means of reducing the voltage/current. I would be tempted to cut the wire in half, though, to get 1400W.

5) At lower power levels, it can take hours to reach temperature if you ever do. Higher power can waste less energy since you have more heat doing useful work and less trying to heat the world.

The wire will not heat the work to a temperature hotter than the wire. However, the wire will get hotter if you keep more heat in.

The caution against adding insulation to a commercial kiln doesn't really apply. The commercial kiln already gets as hot as it is supposed to and adding more insulation can upset that balance. However, with sufficient insulation, you may need to worry about controlling the temperature of the heating element. Unmelted metal in the furnace will help to keep the temperature down; once it all melts, temperature will start to rise faster.

You might be able to use a piece of nichrome wire and an ohmeter to measure the temperature inside (provided your connections don't deteriorate with the heat). Monitoring the current through the heating element itself vs the voltage can also give you an idea of the temperature of the heating element itself.

700W is not much. It is pretty easy to lose that much heat.
Here is someones electric foundry: looks like more insulation, 2000W, and covered:
Gets hot enough that overheating is an issue.

http://www.abymc.com/Articles/Forums/Electric.html
http://www.heatersplus.com/nichrome.htm
http://www.wiretron.com/nicrdat.html

Jpfalt
12-28-2009, 01:32 PM
I took a look at the specifications on supplier sites for the wire and the refractory.

The wire should be good to 2190 deg F. and the refractory will be good if you used the Kastolite 23 or 25.

For the volume of the furnace, your power level should be fine.

However, the insulating capability of your refractory is limiting your final temperature.

I would reccommend that you recast the liner with Kastolite with at least 1/2 the volume in vermiculite added to improve the insulating capacity of the refractory.

Someone recommended Kaowool, which is a very good suggestion if you have some available. In that case I would cast a cavity 3" inside diameter by 6" long with about 3/4" wall and then wrap with kaowool to fit the inside of the paint can. The Kastolite gives a fairly durable liner and the Kaowool holds in the heat.

vincemulhollon
12-28-2009, 01:49 PM
I don't think adding insulation around the outside of the furnace is a good idea. Sure, it'll raise the inside temperature, but it'll also raise the temperature of the outside of the furnace. It may get hot enough to melt the insulation on the wiring, or burn the outside of the furnace itself.

Nichrome melts around 2500.

Lets say the wires run around 2300, the inside runs around 1800 and he wants around 2000 to melt gold. Now you'd like a bit of overheat so the metal doesn't instantly solidify... lets say we'd like to run 2100.

So, you need enough insulation to go from 1800 to 2100, which is 300 worth of insulation. Problem is the heating coils will now gain "about" 300 degrees was 2300 now 2600. Ooops, the heating coils are now a puddle of nichrome.

Can use different materials, but it all boils down to the same problem.

Teenage_Machinist
12-29-2009, 02:33 PM
the refractory is Kastolite 30. The wire is Kanthal, 18 gage. Wound around a 1/4 inch rod (Wheee! That was fun!) and it would be difficult to change the number of coils. I could convert to 220 volt, which would double the power to 1400 watt, although current carrying capacity might be an issue.

When on, the coils glow very orange, so much light in the thing that I can hardly see what's glowing and what is not. A small cap head screw got dull red, but that was measured a few minutes after turn-off, when coils had cooled enough to see anything else glowing (I use welding goggles when it's on full, bright red glow can hurt my eyes) and I have a thermocouple and a PID that will eventually be the controller.

I can get Kaowool cheap. Your idea of putting it in a larger bucket with kaowool is a good one. Maybe building fiberglass would also work as although the outside gets too hot to touch it does not burn stuff really.

The bottom is insulated. About 1.5 inches Kastolite on all sides.

Most of the pottery kilns at school have 1 layer soft brick. I think you are probably right.

rdfeil
12-30-2009, 12:23 AM
TM,

I do not know a lot about high temp furnaces from a technical standpoint. I have however done a lot of work on the controls for them. One thing I can say is it is obvious to me you need more/better insulation. I have worked on electric melt furnaces that look like you describe yours looking like. IE: bright red heating coils that are painful to look at for any length of time. Now for what I think is wrong... The outside temperature may be very warm but you normally should be able to keep a hand on it for a while without getting burned. Some of the furnaces I have worked on run all day at 2000^F and you can hold a hand on the outer skin of the furnace for 30 seconds or longer easily. If your furnace is to hot to touch then WAY to much of your heat is escaping out the sides.

One other note... If you double the input voltage to your heating element the current will double BUT the wattage (heat) will increase 4 times :eek: (2800 watts in your case).

Good luck and keep us posted on your results.

Robin

Teenage_Machinist
12-30-2009, 01:05 AM
Yeah. Sounds like I better not do that without a pulse width modulating control or something.

I think that putting it into a bucket of Kaowool is good idea. My ceramics teacher has been a source of castoffs and surplus like broken kiln shelves.

rdfeil
12-30-2009, 01:54 AM
TM,

Check out this website:

http://www.wiretron.com/nicrdat.html

They have some interesting info on heating element design. If I read it right your 18 gauge NiCr element wire at about 6 amps will top out at about 900-1000 deg^F. If you double the input voltage to 240 volts the element current will increase to about 12 amps and according to the table the element temperature for a coiled element would be around 2000-2200 deg^F. Depending on the temperature you need you WILL need to up the voltage to increase the element temperature and more insulation to keep it in the furnace. This assumes you will be keeping the same element. You really need to make the element smaller with a higher current to increase the temperature and insulate better to increase effency and reduce the risk of burns from the furnace being so hot on the outside.

Have a look at the website and see what you think.

Robin

edit for typo

914Wilhelm
12-30-2009, 03:15 AM
I could convert to 220 volt, which would double the power to 1400 watt, although current carrying capacity might be an issue.

Doubling the voltage will quadruple the wattage. This will risk burning out your element. Increasing the voltage by sqrt2 (1.414) will double it.

Example:
110 volts and 22 ohms = 110 V/22 R = 5 I (amps) Therefore V*I =110 V*5 I = 550 Watts.

if

220 volts and 22 ohms = 220 V/22R = 10 I (amps) Therefore V*I = 220 V*10 I = 2200 Watts.

darryl
12-30-2009, 04:24 AM
I was wondering when someone would point out the quadrupling of the wattage by doubling the voltage- halving the resistance of the element would only double the wattage of course.

It almost seems as though the temperature can get almost hot enough. Adding a judicious amount of extra insulation might give a reasonably good way to keep the temperature from going too high, although if the heating is so borderline it could take too long to actually reach the desired temperature. You might have a case there where removing just some of the length of element would put you right in the desired temperature zone, then you could adjust the temperature after that by wrapping or unwrapping some external insulation.

Actually, I think just adding the external insulation is going to put you in the ballpark without messing with the element. Then the only drawback with that low power element is going to be a longer time to heat initially.

By the way, whatever the element is, low powered or not, as long as there's an input power to it, the temperature of it will continue to rise unless the heat is being dissipated somehow. If no heat could escape somehow, what would happen when you kept feeding the power into it- I think it was Evan who pointed this out. I haven't checked out that wiretron website, but they must be assuming that heat is being dissipated at some nominal rate to make statements concerning how hot an element will top out at. My guess is it's a test in free air, not a confined space with insulation.

dp
12-30-2009, 12:11 PM
Doesn't the resistance of NiChrome wire go up with heat? If so you can't use a constant resistance in the calculations. It would be closer to constant current as in an incandescent lamp. As such doubling the voltage will only double the power consumed.

914Wilhelm
12-30-2009, 12:42 PM
Doesn't the resistance of NiChrome wire go up with heat? If so you can't use a constant resistance in the calculations. It would be closer to constant current as in an incandescent lamp. As such doubling the voltage will only double the power consumed.

True, the resistance of many conductors versus temperature is fairly linear over fairly limited temperature ranges. Nichrome wire resistance about doubles with doubling Fahrenheit temperature. Thus as your temp goes up, your resistance increases and you will not get a quadrupling of power consumption in a small furnace unless heat loss is so high as to keep the temperature the same. The problem is....final temp of the wire will depend on heat loss of the system and thus resistance will vary and it will be somewhat difficult to predict a final power consumption but it will likely fall somewhere between doubling and quadrupling with voltage. Enough variables (heat loss, ambient heat outside furnace, element run away, constant change of element resistance due to it's own oxidation) that furnace manufacturers take the easy way out and place a thermostat in the system to keep the temps constant.

dp
12-30-2009, 12:57 PM
I wonder what temperature outdoor halogen lamp elements (the cylindrical style) reach safely. I know they get pretty damn hot.

Jpfalt
12-30-2009, 03:31 PM
A few years ago I built two rectangular kilns, one was 6 x 6 x 6 and the second was 6 x 8 x 8.

As the heating element I used .041" diameter stainless steel lock wire. Using a cold resistant measurement, I used 35 feet of wire to get 10 amps with 120VAC. I wound the element on a 1/4" rod and stretched it to 8 turns per inch. When final temperature was reached on the kiln, I measured the current and found it drawing about 9 amps. As a result, I shortened the element by 3 feet to get final current at temperature to read 10 amps.

I used the lockwire as the stainless was rated for 2300 deg F and Kanthal was only rated for 2190. The element cost about 7 cents in material. The last 2 inches of the element were doubled over and twisted tightly so that they would generate much less heat on the ends where the power wires were attached.

The box of the kiln was 2" thick insulating brick, 25# per cubic foot. This was cut with a hacksaw blade and grooves were cut for the element to sit in. I used a 3/8" carbide bit in a router to cut the slots (the 25# per cubic foot insulating brick cuts very easily.

The outside of the kiln was clad in .030 thick aluminum sheet and a rising front door using the 2" insulating brick was fitted.

It takes them both about 30 minutes to stabilize at 1800 deg F. depending on the content of the kiln. I'm using a Newport temperature controller and a variable output solid state relay to switch the power on and off and set the heatup rate.

Teenage_Machinist
12-30-2009, 09:15 PM
Yeah, I'm gonna kaowool the outside.