Tuesday, October 23, 2007

Amazing feeders Part 4

For new readers, this long story starts with Part 1, but I'm too lazy to provide a link.

The physics of feeders is fascinating. Everybody just thinks they are a stupid tube, like on a car exhaust, but they are different, especially when bent. Feeders are extruded as a solid tube, but each must be bent several times, in order to fit in the reactor. The physics comes in the reaction of the metal, as it stretches, shears and compresses to the new shape.

We all know how to bend a soft tube, like a copper pipe (if you took shop in high school). For those who didn't, bending a tube isn't that easy, since you want to maintain the same inside diameter. For example, if you take a cheap straw from McDonalds, and you bend it, you get a crimp where the straw failed. You can't suck up your milkshake with that! You could get one of those fancy bendy straws, but that's cheating!

In order to bend a tube without collapse, you need a proper setup. The most important thing is to confine the sides, so they don't buckle out. Here's a picture.

This is a very soft tube. You can imagine the unbelievable power require to bend the feeders! I find it amazing that at the time, this was considered to be normal tube bending, and no scientific investigations were done on what was really happening.

This picture shows the process for the feeders. As the tube is bent, the outer steel must be thinned and put into a yield state. The inner bend may or may not be yielding. The sides are exposed to a complex stress state, that involves high compression on the outside, and shear on the inside.


After the bend, the tube is in an unknown state. We don't know how much it has thinned or thickened, or what has happened to the material properties. We do know that all steel has a microscopic grain structure, and that deformation, cooling after extrusion, and alloy compostion, all affect this structure. Here is an example.


This is a highly magnified piece of steel, probably with some etching done to enhance the grains. If we sheared this steel, you would probably see most displacement along the grain boundaries, and the grains stretched, but the general structure would be the same. The disturbed grain boundaries would be more susceptible to chemical attack.

At this point, we know that the initial conditions of the feeders are unknown, the general science is unknown, and somebody is now going to make projections on performance twenty years later. If their analysis varies up or down a few percentage points, it means the difference between spending a billion dollars, or not.

-to be continued

No comments: