All --
Thermal expansion data is readily available on the Internet. For brass (nickel-silver is a brass alloy), the coefficient of thermal expansion is .0000104 inches per inch per degree Fahrenheit. To put it in plainer terms (e.g., doing the math), a 30-degree change in temperature will alter the length of a three-foot section of rail by 0.011 inches. A 100-degree change results in .037 inches. If you solder three 3-foot lengths together and your layout is in an uninsulated attic in Minnesota where the yearly attic temperature ranges from -30 to +120, you're looking at 0.168 inches, or about a sixth of an inch. Lest you think this is insignificant, it's not. Even a three-foot rail length that's spiked only every six inches will buckle with only a 15-degree change in temperature if its ends have nowhere to go. Ask me how I know... The reason for this behavior is the modulus of elasticity, which is 16,000,000 psi for brass. Multiplying the thermal expansion coefficient by the modulus of elasticity gives us 166 pounds per square inch for each one-degree change in temperature. The cross-sectional area of a .125 rail is around a hundredth of an inch. Therefore, every five-degree change in temperature sets up about eight pounds of force if the rail ends cannot slide. As an experiment, pick up a three-foot length of .125 rail, place it between the palms of your hands (each end against each palm), and estimate how much force it takes to make it buckle sideways. It's probably in the neighborhood of two to five pounds. Dick Karnes Dick Karnes
