Magnetic field of a MRI...

[AndrewBworth]

No, not really - but if anyone can show me wrong I am interested. I wouldn't worry about pulling in a Buick though. But anyway - a metal object like a car in the earth's magnetic field is full of tiny magnetic vectors that are in disarray. They add up to practically nothing. When they are in a significant magnetic field, all of a sudden, their direction becomes important - going with the magnetic field is lower in energy than going against it. So, now, our system containing the magnetic field and our metal object is lower in energy than if they were apart. Systems seeking a lower energy is a concept in the field known as thermodynamics. Our system can reduce its energy further if the metal object can move into a stronger part of the field - allowing even more of the little vectors to line up. Now to get a system to do the opposite - to have a field which shoots metals with away to reduce energy, is much less likely. Forces that reduce the energy of a system have a tendency to be attractive. So for example, this is why when things cool down - they tend to go from gases, then to liquids, then to solids. Not usually the other way around. So you might be able to build such a device that repels metals - but it's going to cost you energy. When a metal gets pulled into a magnetic field, we can think of this as a relaxation - a "downhill" process. Going the other way, though, and you'll at least have to put energy into our system. One easy way to imagine the energy we'd have to put into the system is to imagine the work it would take to pull a metal object out of the field!