In the last couple of weeks I’ve been fascinated by the amount of thermal expansion demonstrated by the chimney flue in our new house. Like many New Zealand houses, there is a log-burner located in the living area. The flue basically consists of a vertical column of large (steel?) cylinders slotted into each other, going from the burner straight upwards through the ceiling and into the chimney on the roof. We have a two-storey house now, so the flue is pretty tall, and you can see it all the way up.
When cold, the lower section (cylinder) is wobbly. You can slide it up and down a bit – up until it is snug into the cylinder above it, or down until it is snug into the burner below it. By a ‘bit’, I reckon nearly a centimetre. I can only assume that the whole stack of cylinders is well secured at the top (the second and subsequent ones are all screwed into each other), so the thing doesn’t all drop with gravity.
However, with the fire going, this lower section starts to fit snuggly at both top and bottom ends – and when its really hot, there is no play at all on the cylinder. (N.B. I haven’t actually tried holding the thing and moving it when it’s hot – in case you are worried – but I can see it is fitting very snuggly). I can watch and see it happen. It gets longer. Quite fascinating.
Really, there’s nothing terribly surprising about it, but you don’t usually ‘see’ thermal expansion happen before your eyes. (The effects perhaps are more clearly seen – e.g. buckling of rails, cracking of concrete, etc.) Stainless steel has a thermal expansion coefficient of about 2 times 10 to the power minus five per degree C. That means a length of 1 metre will expand by 0.02 millimetres for every degree raised. Raising the temperature by 50 degrees C will lengthen it by a millimetre. I’m not sure exactly what temperature the exhaust gas gets to – but Wikipedia says 600 degrees C for the combustion of wood inside a properly designed burner is possible – and the bright orange glow from blackbody radiation says that it’s pretty warm in there. So a one metre length would stretch by about 12 millimetres, which I reckon is reasonably consistent with what I see. (The flue itself will be rather lower temperature, though).
Overall, it’s a nice little demonstration of some physics. And it keeps the house warm.