Yes, you can't even turn them off by unplugging them, as it were. A certain kind of emergency shutdown dumps the helium needed to allow superconductivity (and thus maintain the magnetic field). It costs like $20,000 to do this and is quite dangerous in itself.

Yeah, I should have thought this through. I knew you kept them cool, and with a bit of thinking it should have been clear that you don't turn a superconducting coil off, since nothing's powering it... ;-)

It's called MRI quenching and dangerous it can be[1].

[1] http://www.youtube.com/watch?v=1R7KsfosV-o

Modern high-field superconducting MRI magnets can hold tens of megajoules of stored magnetic field energy, equivalent to the explosive energy in several kilograms of TNT.

You know how it takes months to power up and down the Large Hadron Collider? MRI machines are kind of like a mini version of that.

and it takes weeks to power them down too, else you get the dreaded quench...

http://en.wikipedia.org/wiki/Superconducting_magnet#Magnet_q...

Depends on the MRI magnet type apparently, but with modern superconducting magnets, yes:

http://health.howstuffworks.com/mri.htm/printable

http://en.wikipedia.org/wiki/Superconducting_magnet#Persiste...

They're not. They're only magnetic when they are in operation. I've been in them, I've been with people who've been in them, and my family is in the medical field. If they were fully energized all the time, you wouldn't be able to bring people in and out of the room on a stretcher.

I'm not an MRI tech but I work on MRI image analysis academically, so I know a bit more than usual by hearsay and I see them used all the time. Why are there "MRI magnet is always on" signs in the medical imaging areas of the med school here? Are you sure they don't use titanium stretchers near the field? Why do they perpetually supply the MRI with liquid helium?

My understanding is that even when they are not used there is a field of between 1 and 3 teslas, depending on the scanner. (A quick google search for things like "mri magnet is always on" and "mri safety" and such seems to confirm this.)

Edit: Also, you can get closish to the scanner with a ferromagnetic object without a problem. The field doesn't encompass the whole room or anything. 10 or 15 feet away will be fine.

> The field doesn't encompass the whole room or anything.

Well, technically it encompasses the entire universe - or least the part that is in the light cone since the field was energized. :)

But magnetic fields drop off by distance^3 or ^5, so the strength drops rapidly with distance.

What do you mean ^3 or ^5? Is there some funky physics I'm misunderstanding here?

Because magnets have two poles, the force drops off by varying amounts, depending on the shape and size of the magnet. (The closer pole attracts, while the farther one repels. The other magnet also has two poles, which makes things even more complicated. For example in a short magnet the second pole has a greater influence than it does in a long one.)

It also depends on if you are interacting with another magnet, or with unmagnetized iron. ^3 or ^5 is just an approximation - it can go to ^7, and it's not a definite number, it varies.

See: http://www.exo.net/~pauld/activities/magnetism/forcebetweenm...

>> The field doesn't encompass the whole room or anything.

>Well, technically it encompasses the entire universe - or least the part that is in the light cone since the field was energized. :)

As permittivity is generally a function of frequency doesn't this mean that the progress of a magnetic field through space can differ to that of a [theoretical] light cone (which bounds the volume of causal connectedness)?

Magnetism gives me headaches.

I apologize, I should clarify my point.

They're not magnetic enough to pull a chair off the floor or things across the room (as the story indicates) when they are not in operation. You can walk around the room with your watch on (for example), but if they fire it up and you're in there with it, it's destroyed.

Just look how close they have to get the steel oxygen bottle to the center (in the video linked above) before it moves. It's almost inside it already.

There are low field scanners that use conventional electromagnets, which aren't always on. However, the scanner shown is a 1.5 Tesla (T) or 3T scanner (difficult to say which from just the housing), and the main magnetic field is always on. The main field is supplied by a superconducting ring immersed in liquid helium. As a previous poster noted, the only way to shut one down is to dump the helium, and the cost to bring the scanner back up is in the tens of thousands of dollars. There are smaller electromagnets involved to apply field gradients across the bore of the scanner (on the order of 50 mT/m) which are off when the scanner is not operating, but these are quite small compared to the main field.

My Ph.D. dissertation was on data acquisition and reconstruction techniques for MRI. I've logged hundreds of hours operating high field scanners like the one shown, and dozens of hours being scanned for various research studies. I've also (carefully) hauled a variety of strange things in and out of scanner rooms. Most of the tools we used were non-ferrous, and we had to be extremely careful with the few ferrous pieces of equipment we had to use. The magnetic field does drop off rapidly as you move away from the scanner, so objects more than 10 feet or so away are unlikely to be pulled in.

Sorry, but you're wrong. The main magnetic field is always on. It's produced by superconducting rings that are "ramped up" when the magnet is initially installed and remains on at all times, even when the operator console is powered off. Basically, the ramp up procedure sets up a current loop in the superconducting coil via induction in a controlled manner. After the initial ramp up, the field can only be shut off by bringing in an engineer and special equipment to perform a ramp down procedure, the operator executing an emergency quench, or an accidental quench occuring due to a failure of the refrigeration system. Basically, a quench happens if the temperature of the superconductive rings exceeds the temperature threshold required to maintain superconductivity, the resistivity of the ring material becomes non-zero and runaway heating occurs because of the high current in the ring (lots of amperes), the runaway heating of the ring causes the surrounding liquid helium to go supercritical and rapid (and potentially explosive) boil off occurs.

The electromagnets which are only turned on during operation (used for the gradient fields and for shimming the main field) are insignificant in magnitude compared with the main field. The main field is powerful enough to lift a ferromagnetic chair off the floor if the chair gets close enough to the bore without the gradients being active.

Sorry, but you're (mostly) incorrect. The field for all modern MRI machines is provided by a superconducting magnet, this magnet is "on" all the time, by virtue of the superconducting ring current. The additional fields applied during analysis are inconsequential.

Where you are right is that the fields decay quite quickly (exponentially, in fact) as you move away from the magnet. For the metal chair in the picture to be "sucked in" to the field, I would assume that someone would have had to basically insert it into the bore.

Are you sure about the exponential decay? My first guess would be something closer to inverse quadratic, but that's probably wrong.

It's 1/r^3 for a dipole magnet.

Equally in all directions?

I'm quite sure you are wrong.

You can use a fiberglass, or aluminum stretcher.