60Hz induction heating help

[bill_c]

I want to use induction to heat a steel pipe that has a steel bar inside.  I want to keep this as simple as possible, like just fuse, wire, and a relay.  This will only be used for about 10 minutes a day at startup so I'm not worried about high efficiency as long as I don't damage the wire.  I will be using a thermocouple or RTD to measure and control the temperature.
I did a test, wire got warm but you could still hold your hand on it, pipe got hot enough in a reasonable time (about 300F+ in about 15 minutes), here is the setup:
1" steel pipe (1" ID, 1.3" OD), 25" long with steel bar inside (actually 40" long but I don't want to heat all of it)
25V AC, 20A, 60Hz   (this adds a transformer that I would like to avoid)
About 200 turns of #12 solid THHN wire wrapped around a 2.375" OD PVC sleeve. (just because that is what I had on the shelf)
450F max temp operating of pipe, pipe will be insulated.

I want to stay with the 60Hz for simplicity but I would like to avoid a transformer if easily possible. Steel pipe size will most likely be a larger diameter (less than double) and or a longer length in final design. I will use proper magnet wire and hope to keep it outside the insulation that is around the pipe.
Questions:
1.  Would it be possible to use a different size wire and more wraps to be able to run this with 120V or 240V AC instead?  If so what would be best or how do I calculate this?
2.  If the 120/240V is not possible, then would it be possible to make this more efficient with a different wire size or turns ratio?
3.  Does the diameter of the coil make much difference?
4.  Would a thicker wall pipe help or not?
5.  Splitting this into 2x 12" coils would be nice, but how would that change the answers to 1 and 2 above?
6.  If switching this with SSR, should I use zero crossing or random since this is kinda like a transformer with a shorted secondary?
7.  Is there something else I am missing here?

Thanks.

[Vovk_Z]

We use high frequencies (at least tens of kiloherz and higher) because of a strong reason. It is hard to make strong enough Eddie currents at 60 Hz. But if your setup works - then ok.

[ejeffrey]

Induction heating of steel is much easier because of the high permeability and hysteresis so 60 Hz can work.  For non-magnetic materials you really want higher frequency.

If you are happy with your performance but would rather avoid the transformer then you can just use more turns.  For a given magnetic flux there is a constant volts/turn -- so to increase the working voltage 5x you need 5x the number of turns to about 1000.  The current will drop by the same ~5x so you can reduce the wire to ~AWG18 to keep approximately the same restive losses.  This is assuming you do the winding over the same length.  You will need more winding layers to get equivalent performance.

[IanB]

Out of curiosity, if the AC were rectified it would double the frequency with the loss of a sinusoidal waveform. Would this make induction heating more or less effective than with the original AC?

[T3sl4co1l]

Out of curiosity, if the AC were rectified it would double the frequency with the loss of a sinusoidal waveform. Would this make induction heating more or less effective than with the original AC?

Less, because you have a DC component, and it can't go nowhere, you have to sink it through a resistance.  You get doubled AC and harmonics, with all the efficiency of a "class A" amplifier.

In principle you can tap it off with a nice big filter choke, and run a grid-tie inverter to return it to the mains... but, y'know?

Tim

[bill_c]

Thanks guys, going to give it a try with 240V since the machine is wired for 3 phase 240V and doesn't have a neutral conductor from breaker panel (should have checked that first). will try ~1900 turns of #20 wire.  If this was more demanding, I would go for the higher frequency, but this meets all my needs.  It may be a month or 10 before I get to a complete working prototype of the machine, will post results then. 
Any input on which switching is better for less noise here, zero crossing or random?