microwave muffle furnace

[durable_oreo]

I stumbled on a guy who uses a appliance type microwave oven to melt metal (link below). He makes a kaowool enclosure for home-made silicon-carbide crucibles. The kaowool absorbs little in the "microwave" bands and silicon carbide absorbs strongly.

https://www.youtube.com/watch?v=P1VmIYheuU4&t=293s

How hard would it be to design a waveguide that is optimized to dump most of the magnatron's power at a specific point in space? The kitchen appliance must accommodate many different payloads but if you are melting metal, the crucible is always in the same place. There are no constraints on the shape "cooking" chamber as long as there is space for a large cylinder of kaowool with the payload in the center. Blocking radient heat from damaging the magnatron would be an important design consideration. The use of multiple magnatrons would also be of interest.

BTW, there's some nice commercial products out there (CEM) but they are not optimized for a single use. Also, they appear to use an alumina firebrick lining, which seems like it would be an unnecessary parasitic load. If my money tree was more vibrant, I would throw 16k at this question and direct my assistant Jeeves to conduct some experiments.

I do have a machine shop at my disposal, though. Welding and sheet metal are also accessible. The only wave guides I've designed were following formula in a handbook, back in college. At this point, I'm mainly wondering if this would be a challenging problem or if it's feasible.

 

[durable_oreo]

I don't know if the lack of response corresponds to the level of traffic or the quality of the question. Let me ask in another way:

Given that a waveguide for 2.4 GHz is something like 24 x 12 inchs and an 800 W magnatron is the power source at a right angle 1/4 wavelength from a backshort. There are no other ports. A small silicon carbide crucible is suspended in the center near the top of the waveguide, holding a sample (let's say aluminum). How long would the waveguide need to be to have a stable maximum where I can place the sample? Is it as simple as 1.5 wavelengths?

Could 2 magnatron be used? Would a magic T with the delta port grounded be needed or could I have 2 right angle ports, one in the top and the other in the bottom of the waveguide?

In open to any and all input.

 

[durable_oreo]

Oh, the waveguide dimensions are 3.4 x 1.7. That's close to what I have seen in torn down microwave ovens. So my question should be: how to build a resonant chamber with the smallest number of hot spots. I think the answer is going to require simulation. Well, this concludes my idiot microwave blog series. Thanks for calling

 

[IRstuff]

how to build a resonant chamber with the smallest number of hot spots.

These seem to be contradictory requirements. If you're trying to heat a specific thing, why would you care whether there's energy in the corners of the chamber?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[durable_oreo]

That's a good point. A "hot-spot" in some un-occupied section of the microwave isn't wasting energy. There is no load at that location so there are no losses beyond the usual path losses due to distance and a non-vacuum medium, right?

Perhaps I should be asking about power transfer. In a cavity with a standing wave, how much RF energy can be absorbed from one of the points of constructive interference? Using a 1-D acoustic example of a guitar string, it seems like most of the energy could be removed from any point that is not a node.

If a large fraction of the input power is delivered to a load placed at any "hot spot" in the microwave oven camber, there may not be any advantage to "focusing" the microwave power on a certain point in space. As long as my crucible is not too small, it is likely to intersect with several local maxima.

Finally, the trivial case works in my favor!

 

[Comcokid]

The location of any "hot spots" will vary with the location and size of the load (crucible) each time you place it in the microwave as the load affects the standing wave in the chamber. Food does not conduct heat well. Food is turned in microwaves so the food cooks evenly despite hot spots. Your crucible will hold silicon carbide which conducts heat well, so I think any hot spot will have the heat conducted quickly into the whole of the silicon carbide.