re latest announcement, page-76

We did wonder where you were, Geojac, when the fun started a week ago.

Most steels and their common alloys have a coeff of linear exp of 12 x 10^-6 /C although some special low expansion metals are available (e.g. invar, NiFe alloy 0.9 x 10^-6 /C - hideously costly). I don't know the value for well casing.

Supposing particular temperature gradients for the flowing brine and also for the rock strata it is possible to interpolate the temperature differences at increasing depths. (At 4200 m the difference is zero.) My calculation gives a value of 4.3 m for a free incremental expansion with much of the contribution being in the upper 1000 m. It seems a colossal figure. Where would the stress be accommodated? Theoretically all the way down the well wall. What are the adherence properties of concrete to steel vis a vis concrete to rock face? At each start/ stop thermal cycle the well casing tries to shear up and down the well by this amount.

Railway lines sometimes do buckle. In the 45 C heat in Melbourne this summer (unstressed) employees were hosing down the lines.

Juke