Induction furnace design

You call 100 Ton Daily Small scale!? I wouldn't even think of 100 Ton yearly as small scale...

Induction furnaces heat by inducing electric currents in the objects being heated. Since glass is an insulator, no current, therefore no heat. You could heat the molds if they are metal, but not the glass.

Try these sites for ideas

http://www.ameritherm.com/aboutinduction.php

http://www.ameritherm.com/overview_susceptorheat.html

on the design configuration of the furnace.

The sizing of the furnace, of course, is determined by the production volume and the number of runs you want to make.

Then use the microprocessors to control the induction design, moulding progression, and any dynamic heating-policy you wish to impose on the heater.

I did some contract work at U-Carr (Union Carbide) in Cleveland some years back on their electric furnaces. These furnaces where made to melt/recycle 25+tons a cycle of all kinds of metal, and am sure they would melt glass in no time. The basic dims were about 30 ft diam X 14 ft high. The dome contained 3 heating elements/electrodes. I don't have time to do the research, but if you look for "U-Carr Electric Furnaces", it may shed some insight. Instead of reinventing the wheel, it may be more cost effective to purchase one of these. Good Luck.

I have the feeling that Shekhar is interested in becoming knowledgeable in the furnace design science/art. At least that is how I read his posting because of his remarks about the decision - seemingly prevailing already - about the use of microprocessors to control the furnace.

Besides, the induction heating will give more focused localized heating that dispersive heating provided by heat-conduction from heating-elements, and therefore more energy efficient. However, I am basing this assessment on the presumption that the primary emf power-use inefficiency due to the magnetic reluctance that obtains because of the air in the furnace is less that the energy losses due to the "thermal reluctance (?)" of the thermal conductivity of the air in the furnace.

I would think it would be more efficient to melt 100 tons of glass one time and control the flow into the molds (via microprocessing ) versus melting .0001 ton X 1000000 times. I am sure I have some calculations floating around to show that increase mass = increased efficiency in regards to thermodynamics. I guess it would be important to know what the source (25 tonn?) is and what the desired output 1 piece of 25 ton, or 25 pieces of 1 ton, etc..) would be to really be of use to this gentleman. As for the Arts and Sciences perspective, budgeting and funding allocation unfortunately has become more prominent than design engineering these days. Bless those that still have the open bank accounts in the R&D field.

Well, instead of waiting for Shehkar to explain, let us learn from each other and thereby also get Shehkar to learn the knowledge we impart to each other. So let us synchronize our furnace concepts and perform concept analysis; and I will start.

Now, because induction heating provides heat by secondary electric power and hysterisis, the glass which is non-magnetic could not generate electric current by induction. Hence, the glass has to be placed in a melt vessel that must be magnetic. So when the primary emf alternating current flips the magnetic fields in one direction and then the other, the melt vessel generates the electric power and the hysterisis that causes the heating. So in this sense the heat is generated and applied to the glass in a completely localized state - by the melt vessel containing the glass. Of course, the only inefficiency in the use of the primary emf power is the limitation on the magnetic field that gets to be generated due to the magnetic reluctance of the air, and consequentially the commensurate reduction of the electric power created in the melt vessel.

Now on the flip side, when electrode - resistance based - heating is used, the heat has to be conducted from the top of the furnace through the air to the melt vessel which will then get heated to melt the glass. The energy losses accounting would consists of the losses due to the energy used up in heating the air, etc.

Again I am assuming that the design/operation basis on which I am thinking about the application of the heat are as above for your concept as well. Then, of course, I could be mistaken.

I think there would be a break-even point of efficiency between the two methods of melting glass: Magnetic induction current applied to a vessel versus heating air like an oven. Perhaps a fundamental issue may include what is the best way of heating air as another variable into the equation. I would think that would be based on geographics, as in the US gas is currently cheaper than electricity. IF I had to melt one 100-ton block of glass, I would look into a blast furnace via gas. Hence, I will concede that electric heating of air is not the most efficient method versus gas, and therefore will rule out electric air heating for sake of our friend. Let's make some assumptions based on our friend's original post:

1. Glass slaps come in as 1-pound billets.
2. Electricity = Gas in terms of $/joule
3. Manufacturing facilities can handle 1-pound product.

I can see a line of 10-20 electrical heating vessels from either one of two methods.

1. The vessel is heated directly by direct current or
2. The vessel is placed into an electromagnetic field.

Now the issue that comes to mind is material handling. I am assuming this task will be electronically controller via conveyors and PLC like a Bosch system. If is better to keep the melting vessel static, then drop loaded into a mold on a conveyor, then a coil heated container would seem appealing.

When I was at Picker NMR, we made these massive magnetic coils that had to be cooled via liquid nitrogen. They kicked out a lot of heat. Also, everything on the fab side had to made of non-magnetic stainless steel. The rooms where lined with lead to reduce RF emissions, and if one wore a watch in the room, it was fun to watch it get ripped off of someone's wrist (well, if we knew the person). There is that added cost to allow this concept to be user friendly.

Unfortunately, the EMF approach generated heat at the coil itself, and not in the actual fields. I imagine if frequency switch was high enough, the fields themselves could excite the glass molecular structure to make it a viscous state. I wish I had the time and money to explore this option.

As you know, whichever method is used, the heat still would need to contained and somewhat directional. Insulation and compounds will be key for the final efficiency factor.

I have fallen out of R&D field some years back, but I find myself hanging around CR4 to make me feel young again.

Let me add that EMF frequencies can be from either electrical switching at the coil level OR mechanically moving the coils. This made me think of RF heating. Why not? This can be directional and perhaps more controllable and possibly more efficient. Hmmmmmm...

Had some experiences with electric induction furnace to melt aluminum. I have also used them to melt tin and keep it molten for the floating of molten glass on it to make float glass. But, as many have pointed out the glass will not be heated by the electrical induction unless it is conductive of current. Heating a metal and using it to transmit the heat is a viable way to heat the glass.

Question is the glass in the finish mix or is smelting also part of the process he will be doing? So many questions and much more needed in the way of his final use to make the decision's as to what information and direction to point the gentleman.

We are currently using gas to heat our glass and make the compound we use in making a glass insulator. We smelt the compounds into glass using a gas furnace.