If you need to get caught up, check out Part 1.
In this experiment, I will be using the plastic Ork Nobz box set which have had magnets super glued into them using Duro super glue.
Method 1: Rubbing Alcohol
I poured some rubbing alcohol into a old glass baby-food jar that I sometimes use to clean my brushes as I paint. Not sure of the alcohol's affect on the plastic, I broke off a small bit of extra sprue and left it sitting in the liquid for about an hour. It is highly suggested that you cover the container, to keep fumes from spreading. I don't have a lid for this jar, so I used saran wrap and a rubber band.
The test bit of sprue did not seem to have any sort of deformation or softening as a result of being in the alcohol, so I went ahead and dropped in the two arms and Nob torso (already glued to a base) into which the magnets had been glued backwards from the others. The jar was then left to sit out of reach of little hands for about 24 hours.
The next evening, took the bits out of the jar with a pair of tweezers, and dabbed them with a paper towel. I was shocked to discover one of the magnets from the torso just fell right out! The two magnets in the arms came out, but required some work with my Xacto knife. I slid the blade into the narrow gaps around the magnet and wiggled it a bit. Eventually, they too came free.
Unfortunately, the other magnet in the torso was still glued firmly in place. There was no room around this last remaining magnet to slide the blade in without damaging or cutting away at the plastic around it. Since I needed the area to stay intact in order to re-glue the magnet in properly, I opted to move on to the next method.
Result: 3 of 4 magnets removed
Why (I think) it worked
Assuming there were pockets of moisture trapped in and around the super glue bond holding the magnet to the plastic, the alcohol help remove that trapped moisture, leaving behind empty pores in the super glue - weakening it. Thus, less tension was required to break the bond. (I totally pulled that out of my rear, but it sounds good . . . right?)
Method 2: Freezing
Taking the torso into the kitchen, I ran it under the faucet for a couple of minutes, try to ensure that water got into every nook and cranny around the magnet. The model was then placed in the freezer and left for 48 hours - Not because I thought the extra time would matter, I just forgot about it for a day.
Taking the model out, I worked at it a little bit with the X-acto knife and also tried holding other magnets on either side of the embedded magnet. Unfortunately, all I managed to do was hack up the area around the magnet a little and leave a scratch on the magnet itself.
Result: 0 magnets removed
Why it was supposed to work
I actually read this on the messageboard (sorry, don't have the link) where I learned about this method - again, this assume moisture is trapped in an around the super glue bond: When water freezes, it expands. The expanding moisture in and around the super glue creates tiny cracks in the bond, weakening it and allowing the bond to be easily broken.
Why it didn't work
Not sure really. Perhaps the alcohol effectively removed what moisture may have been trapped, so there was nothing to expand upon freezing. Perhaps the magnet is set in such a way that there is no space for either liquid - alcohol or water - to penetrate.
According to Loctite's poorly designed website (the makers of the super glue I was using), soaking the bond in warm water might work. There also exists super glue remover, but I'm really only interested in cheap, non-hazardous alternatives.
I would also like to get a new digital camera - one with a working LCD screen, exports directly to my laptop, and doesn't try to cram a ton of noise-producing megapixels through an undersized processor - but that's another rant-er . . . story for another time.