Why is Microwaved mac & cheese burnt where they touch? 
After reheating cold about 1.5 oz. of Annie's Mac & Cheese shells for 15 seconds on high power in the microwave, the mac & cheese was burnt black only at certain points where the pasta is touching each other. Does anyone have an idea of what might be going on?
 A: I get the same thing reheating some discs of glazed carrots.  And there are several videos of folks doing this intentionally with grapes.
An article published last year in PNAS says this will happen with almost any pair of similarly-sized object with sufficient water.  The shape of the pairs appears to set up a resonance that concentrates the electric field at the point where they touch.  The higher intensity field there heats the surroundings and burns the food.
A: I'm "fairly confident" that the more esoteric explanations are overly so. 
Voltage is induced in closed  conducting paths in an RF field (here a complex E-M field with various nodes due to the cavity but that is not a major factor). 
AT points of contact the resistance is high and the flowing current creates "i squared R" heating. (Power dissipation = current squared x resistance). This does not requires resonance, appropriate wavelength items or other RF interacting factors. 
For a spectacular and odiferously vile effect


*

*Take an OLD and UNWANTED CD or DVD
The CD / DVD will be utterly destroyed in the process. 

*Stand it on a drinking glass or similar in the microwave oven.

*Set microwave to 3 to 5 seconds (longer is a mistake)

*Go

*Superb firestorm in miniature with "sheet lightning across the CD surface.

*Do not breathe the fumes. Conclude that it was fun but should have been dome outdoors. Wow!
A: @BowlOfRed hit on it solidly. The noodles are acting as waveguides, because of their size and shape. Where they meet, a contiguous surface is created, but with a much higher resistivity, as it's a narrow point contact. 
Based on a table of refractive indexes and a table of frequency/wavelength, this effect would be especially effective when the total length of the joined area is at a resonance of the base wavelength of 3.54 inches (2.5ghz in air is 12cm, water is more refractive). So if they match in physical dimensions to 3.54 inches, 1.77 inches or 0.885 inches, the effect will channel a decent amount of energy. If that boundary between globules has a high resistance and carries a high amount of energy, it'll dissipate a lot of that energy as heat and light, potentially even creating plasma flashes as in the ideal natural example, the grapes mentioned by @BowlOfRed. Google "Grape Plasma," or just check out this video.
P.S. I don't know how big the Auntie Anne's shells are, but the curved concave shape is probably increasing their actual resonant size by something like 1.5 times. I'm not exactly sure how to work the formula on an irregular or non-ideal shape. Does that sound more like a match?
P.P.S. The YouTube dude captured some awesome shots of arrested plasma. He even talks some of the actual science of it. Trés magnifique!
