OT: inductive stoves & eddy currents

[capslock]

Off topic - but still involving effects like magnetism, hysteresis, eddy currents, skin effect.

We will shortly be moving to a new home. The great thing is that it has some 700 square feet of really nice basement rooms, so I can store all my drivers, tools and electronics, and actually work on a project in one go. Before I can get to that, there are some other things to take care of which my wife insists have higher priority, such as selecting a kitchen. So we have been looking at appliances. Even though inductive stoves have been around for a long time, it seems the marketing departments have now decided to make it a mainstream technology.

One thing all manufacturers stress is that the inductive heaters will only work if the bottom of the pan is magnetic, and they suggest that one checks whether a magnet sticks before using the pot or deciding to buy a new one.

Apparently, those heaters contain a coil that is driven by a 20 to 30 kHz AC current. This will induce a magnetic and electric field in the iron of the pot and heat it because of hysteresis (magnetization) losses. Also, as even cast iron has some electric conductivity, there will be eddy currents and losses due to the electric resistance of the material.

Now, I would assume that the AC field would also induce a substantial eddy current in aluminum or copper, heating those materials through resistive losses. Why would this not work? I assume the skin effect would still be negligable at those frequencies. So would this short the coil?

The only thing that I would see not working is some alloys of stainless steel which are non-magnetic and have a pretty low electrive conductivity.

[Feyz]

Don't know anything about those ovens, but one thing that comes up to my mind is that they may be needing the iron or steel pans so that a good magnetic coupling will occur.

[john k...]

(JPK) There are two sources of heating. One is the generation of the eddy currents. The second is due to the rapidly inverting magnetic fields in the material being heated. The rapid inversion of the fields in the material being heated results in significant losses which result in the second mechanism of heat generation. Obviously this is absent in non-magnetic materials and the only heating mechanism is eddy currents. Thus there is a loss in efficiency if the pans are not steel or iron. To improve efficiency of heating non-magnetic materials inductively the AC frequency must be increased.

[capslock]

I posted the same question on a German forum, and one engineer answered who designs inductive heaters for industrial applications. They use 100+ kHz. He said that in his applications, it was a question of impedance matching. His generators are usually matched for ferromagnetic materials or even austenite (non-magnetic stainless steel). For copper he'd have to redesign the LC coupling circuit.

But he said he had no idea if this would also be true for 20 - 30 kHz.

Another interesting thing that came to my mind since this morning:

The marketing tag line is that induction is faster than anything else, even halogen and gas. They have some numbers of how long it takes to heat 2 l of water, and inductions comes out as 20% faster than gas.

The argument is that induction heats the bulk of the pan's bottom, whereas everything else heats the bottom of the bottom...


Now think of this: if those stoves only work with materials which have a pretty poor electric conductivity (cast iron, some stainless steels), this means that those materials have a pretty low thermal conductivity, too (those two are roughly proportional).

Consider a cast iron pot with a 10 mm bottom. With induction, the heat has to travel an average of 5 mm. With gas, halogen or Ceran the heat has to travel 10 mm, but if one chose a pot with a copper bottom, the thermal conductivity is roughly 40x higher. So with the proper pot, pretty much anything becomes faster than induction!

[john k...]

(JPK) Except with that with gas you have a lot of heat lost to the surroundings so only a portion of the heat of the flame actually goes into heting the pot. Also, heat transfer to the pot from the flame is via. convection, etc. I wouinder more about the relative efficiency of the systems. How much energy does it take to boild 16 oz of water?

As for the response you received from the German source, it sounds reasonable. Rasing the frequency means the eddy currents flow through a smalller cross section of the pan, near the surface, due skin effects. So there is greater resistance, higher I^2R heating, etc.

[capslock]

Well, energy being more expensive here (as it is taxed by CO2 output or other environmental toll), I have those figures. Natural gas is about 4x more efficient than our average mix of electrical energy (70% fossil, 5% renewable, 25% nuclear fission). This is due both to conversion efficiency and distributive losses. Using gas instead of electricity saves about $200 per year for an average household (but would for us involve getting a new line put in from the basement into the kitchen).

So if you have about 30% convective losses, gas would still be much more efficient. As for speed, a typicial gas burner has 3.5 to 4.5 kW compared to 1.5 to 2.2 kW for an electric heater, so even with 30% losses it would still be faster.

The traditional electric heaters in which the resistive element is covered by a steel plate are very slow, because of their lower power and conduction time. Because of their large thermal capacity, you also loose energy once you're finished.

Going back from gas to one of those for a time was an intellectual challenge, because I was not used to the delays any longer. Ceran is a reasonable compromise, but still noticably slower than gas.

[Feyz]

Instant metal melting

Thought others might find this interesting too, video of metal boiling within seconds:

http://www.deutsches-museum.de/mum/video/e_vid10.htm

[john k...]

(JPK) No argument that gas (or open flame) is generally considered the best for cooking. You don't see too many (if any) electric (resistance or inductive) stoves around here in commercial establishments.

[Feyz]

A counter top cooker, or a closed oven?

Is this a counter top cooking appliance, like a range, or a closed oven? If it is the former, I still think it is about getting the magnetic coupling. In a closed oven, you can run the coil loop all around the pot that is being cooked, which will be subjected to allmost all the magnetic flux generated by the coil. But if it is a counter top heater, the pot needs to become part of the magnetic circuit so that enough magnetic flux will pass through it. I think this is done by putting coil or coils under the cooking surface, which are magnetically short circuited at the bottom with a bottom plate, similar to a bottom plate of a speaker motor. The pot that's put on the heating surface becomes the top plate that conducts the magnetic flux between the coils, again similar to the top plate of a speaker motor. If the pot is not ferro magnetic, a smaller amount of flux will pass in it, most of the flux will be stray field outside. If it was possible to wrap the pot within the coil , like in the industrial ovens, or a closed oven, one wouldn't need any ferromagnetic property, because the element to be heated is inserted into most of the flux that has been generated in that case. Just my thoughts, as I said I don't know much about them anyway...

[capslock]

These are counter top appliances, not ovens.

[whoaru99]

Depends on what you consider better.

Gas stoves are usually more controllable for temperature variations when cooking. However, I've seen in at least one magazine that tests consumer products that the resistive element stove is usually the fastest in a time-to-boil test.

[JSG]

For those who care...

In most cases, eddy currents are the dominant effect for induction cooking. Cooking vessels with high permeability have a tremendous advantage of non-magnetic materials. The skin depth is also inversely proportional to sqrt(mu), which makes the effective resistance of the magnetic cooking vessel much higher than with non-magnetic materials. Typical mu values are several hundred or higher. The pot is the secondary and load of a transforemr, and higher resistance means a lot less current is required to get the same heating. Less current in the pot means fewer amp-turns, less voltage drop across the leakage inductance separating the coil and pot, less loss in the coil, and in general makes life a LOT easier. The substantially higher resistivity of magnetic materials also helps a little, as the resistance will increase proportional to resistivity, but the skin depth goes down as sqrt(resistivity), the net effect being an increase in resistance with sqrt(resistivity). All this assumes the pot is several skin depths thick.

Also, consider this: the coil induces eddy currents in itself as well as the load. If the coil and load have similar conductivity, comparable amounts of heat will be generated in the coil. This is unworkable for a typical consumer appliance with a lot of electroincs in it. Things like litz wire can help, but litz is not a panacea. It is a lot cheaper and easier to just use a nichrome coil at 50/60Hz (not nearly as interesting, though).

One point is that with inductive heating, you can pump a lot of power into a load if you want. You are not limited by the thermal limitations of the cooktop material. In addition, it can be a lot faster response because the heat stored in stuff surrounding the pot is a lot less than with most cooking modalities.


John

Edit: Gas stoves tend to have around 30% or less of the energy in the gas delivered to the pot under typical conditions.

[Dennis H]

I'm no chef but I wouldn't want to be limited to iron/steel pots. Since I usually end up washing the dishes, I have a strong preference for teflon coated aluminum.

[jdybnis]

:B I can see it now, Audiophile cooking ware, Alinco pots and pans.

[Feyz]

.... with kevlar coated cooking surface, kapton handles, copper shorting rings on the base, white ceramic dome lids with ears for controlled gas escape :lol: