"Oh it's dark as a dungeon and dank as the dew, where the dangers are many and the pleasures are few." So goes the old country song about the life of the coal miner. The danger is real as "underground coal miner" is at or near the top of lists of the most dangerous jobs. Roof collapses, in particular, have led to disasters that killed hundreds of miners.
Bolting is one way to stabilize the mine roof. Overhead holes are drilled and long steel bolts are inserted and wedged deep in the rock. Anchor plates are fitted over the protruding end of the bolt and a nut tightened to put the rock under compression. If all goes well the roof stays in place while the coal is being mined.
Although coal seams only a few feet thick may be practical to mine, drilling holes for the bolts may present a problem. A rigid bit long enough to drill the needed depth hole may be longer than the ceiling to floor distance. One solution to this problem is to use a flexible coupling to connect the drill to the bit. Roof bolts are on the order of 1 inch in diameter, so it takes a robust flexible coupling to transmit the needed torque.
Some years ago a Boston-area engineering firm was trying to develop such a flexible coupling. The basic design consisted of two counter-wound approximately 1-inch diameter coaxial steel springs. These were welded on one end to a chuck for the drill bit and the other to an adapter to fit the drill. The spring wires were about 1/2 × 1/4 inch in cross section. The flexible coupling could be of any length, but, in the case at hand, it was about 8 ft long.
Pelloux's law* for metal failure states, "It always breaks where it's welded." This case was no exception and the engineering firm called me in as consulting metallurgist. I sectioned several couplings through the weld region and found, sure enough, there was a whole lot of cracking going on. The reason was not hard to find.
Spring steels typically contain approximately 0.5% carbon, by weight, for strengthening. This is a lot of carbon. Cooling of such a steel after welding tends to lead to formation of martensite, a very hard, brittle phase which cracks easily. The carbon content of the springs was almost too high to allow welding at all.
Cracking may be minimized by appropriate heat treatment before and after welding. Manuals call for about 200C preheat and 600C postheat treatments for a 0.5% carbon steel. The preheat is for stress relief and the post heat tempers the martensite to make it softer and less brittle. My report recommended these heat treatments.
Here enters the human element. The engineering company's welder was a cantankerous so-and-so who didn't hold with all this fancy heat treatment stuff. He also liked to use gasoline for cleaning up, just to show his independence. His attitude reminded me of a number of craftsmen I have met, including my own father. But somehow my report convinced him. I suspect that the Professional Engineers stamp and MIT affiliation on the report did much of the convincing. He started using the recommended heat treatments and the cracking stopped.
*Enunciated by MIT Professor R.M.N. Pelloux