Heating rusted rotor to remove it -- Show #1318

I a quite disappointed in your misunderstanding of the physics associated with thermal expansion. You must have misunderstood, or forgotten, the science that you studied at MIT.

When heating a metal ring the inside diameter of the ring increases. How can it be any other way? If the metal is expanding or lengthening, as it does for any material with positive thermal expansion coefficient (as most common metals exhibit), how can the inner circumference decrease or remain the same. Therefore the ring, in obeying the laws of physics, must get larger. The thickness of the ring itself will get larger too, but this will result in further increase of the outside circumference, not the inside.

Of course the shaft on which the ring is mounted would also increase in diameter, if heated. So the shaft must be kept cooler or made to expand more slowly than the ring.

I have seen this done hundreds of times in machine shops. A bearing, flange or other metal piece with a hole in the center is heated. The shaft or pipe is not heated or frequently cooled with ice or dry ice. If an assembly is being taken apart, this procedure separates rusted, interference fit or swaged fittings handily. For a new assembly, it improves the ease of putting the parts together for swaging or interference fitting.

Sorry guys, your explanation of the reason heating a disc drum works to help in its removal is BOGUS, a word you use often. This is on par with the “two” error you keep broadcasting.

Additionally, I seriously doubt that the heating changes the crytalline structure of the rust to break the bonds between the metals, as you suggest. In fact, it probably helps to reinforce the bond by increasing oxidation rate. Of course this does not override the mechanical fracture stress caused by expansion of the ring hole away from the shaft, as long as the ring hole size is expanding faster than the increase of shaft size caused by the same heating.

Tom and Ray, figment is absolutely right, the inner diameter will increase. Consider the alternative - if the inner diameter of a nut decreased with heat, nuts and bolts would work well only in a narrow temperature range, any colder and the bolt would slip or fall off the threads, any hotter and the nut would become unremovable.

I’m with Figment as well, and so is – of all people – Martin Gardner. From his “Mathematical Carnival” (Knopf, 1975): “Q: A piece of solid iron in the form of a doughnut is heated. Will the diameter of its hole get larger or smaller? A: [The doughnut] keeps its proportions, therefore the hole also gets larger. The principle is at work when an optician removes a lens from a pair of glasses by heating the frame or a housewife heats the lid of a jar to loosen it.” Gardner is on slightly shakier ground with his second sentence (frames get softer as well as bigger when heated), but I think his first sentence is indisputable.

You are all correct. We had a discussion about this a while ago, the last time that show aired, I guess. One way to think of it is heat makes the ring get bigger, like copying a picture of it at 110% on a Xerox machine. So everything gets bigger: the outer diameter, the hole diameter, and the thickness of the ring.

I think the boys from MIT and Andy are just mis-remembering the experiment. Here is the actual:

Another way to think about how this works is to consider the ring to be a wire of infinitesmal diameter. If you cut the wire, straightened it out and heated it, we all agree it would get longer;
now if you join the ends to make a circle, this has to make the ring get larger as you heat it. A
solid ring is just a whole bunch of infinitesmals.

When I heard the sphere through the ring segment I was reminded of the PBS show Woodwright’s Shop w/Roy Underhill (I think). One time he made a wagon wheel. You know with an oak hub and wooden spokes from the hub to a wooden rim. The final step, to make the wheel durable, is to place a metal band around the rim. He constructed the metal band so it wouldn’t fit around the rim at room temperature. Then he built a fire and got some coals going really hot, put the metal band in the fire and let it sit for a while, then using some tongs to pick up the hot band he placeit around the rim and it fit easily. After it cooled and contracted it made for a nice tight wheel.

So the inner diameter of a thin metal band at least does expand with temperature. As @tc says above, the caller must have mis-remembered the experiment he saw.

I got to thinking about what would happen if didn’t have a band, but you had a small hole in a large, thick piece of metal plate. Like if you had a one inch thick plate of steel 1 meter x 1 meter, with a 1 cm diameter hole drilled in the middle of it. Would the hole get bigger or smaller upon heating the plate?

It seems like to me it would get bigger, but I’m not entirely sure.

Yep, bigger. All dimensions increase proportionally.

I guess the only way the hole would get smaller on heating is if the outer surface of the part was contrained in some way so it couldn’t expand in that direction. Then the only way the metal could expand is towards the hole, and it seems like it that situation the hole could get smaller on heating.

Here’s something I didn’t know, related to this thread, found it by accident the other day. There are some materials that have negative thermal expansion coefficients. They contract, rather than expand, upon being heated. Cubic Zirconium Tungstate for example. So it is possible if you use the right material to make a circular shaped band that would contract on heating. Could come in handy for certain applications I expect.


Another substance that contracts when heated (when changing from solid to liquid)-WATER. That happens because of the polarity of the molecules.

Water as a liquid is densest at 4C and expands on either side.

This is separate from what happens when it freezes.