At GE in Schenectady, in an armored vacuum tank, a steel blank for a gas turbine wheel weighing three and a half tons, heated almost to 1,000 degrees, and spinning at six-hundred radians per second (5,700 rpm) was under routine plastic stress relief— by means of centrifugal forces—to redistribute stress left over from cooling after the casting process.
In order to eliminate vibration from small rotational imbalances it was suspended from an eighteen inch long, three-quarter inch diameter steel spindle sufficiently strong to support the weight and flexible enough to allow the wheel to rotate about is center-of-mass rather than about its geometrical center—thus eliminating vibration. Depending upon differences the spindle was, therefore, urged into a slight “S” shape in relation to the vertical.
Unexpectedly, after an hour or so, the spindle broke—releasing more than one trillion Joules of kinetic energy.
The result was catastrophic. Massive wheel fragments tore through the sides of the containment tank and concrete pit burying themselves somewhere in the surrounding earth. Furthermore, it was not the first time this had happened even though no fault could be found in the design of the failed suspension rod.
Jackson and Moreland had been hired to redesign the pit, the tank, and the method of wheel support. I was assigned to the mechanical work.
Gas turbine wheel blanks are spun hot and at speeds high enough to allow inevitable and unwanted internal stresses from the casting and cooling process to “relax”. The temperature is enough slightly to plasticize the steel, and the speed is sufficient to produce the necessary centrifugal forces.
Without further study GE had given up on the failed spindle configuration and opted for a new thrust bearing type of support which was the path that we were directed to pursue. Nevertheless I was curious to know the answer to the spindle failures and set aside some time to investigate.
While thinking about the problem it occurred to me to consider something others had overlooked—the diurnal rotation of the Earth. The massive wheel is, after all, a gyroscope and if its axis of rotation is disturbed in some way resisting forces come into play. And some further calculation showed that such was the case. 
It turned out that at 42.5 degrees north latitude (Schenectady) the wheel’s axis of rotation is, itself, constrained to shift at about ten degrees per hour , sufficient to produce a gyroscopic moment (twisting of the plane of its rotation) to force a steady state bend in the spindle of almost two degrees. Any bend produces stresses which, owing to the rotation are reversed, in this case, at a frequency of 95 cycles per second. Steel samples in rod form are routinely tested for the so-called fatigue strength in specialized machines that load and stress the sample in exactly the same way as in the case of the tilted turbine wheel. The so-called endurance limit of a material is measured in total cycles undergone to failure under a known stress.
The endurance time of the steel used—loaded as in the suspension—was about two hours under the action of the Earth’s rotation alone; including neither the weight of the wheel, nor the vibration accommodating “S” bends.
In this light the failures were not surprising and probably could have been avoided.
 There is no additional effect owing to the Earth’s revolution about the Sun because the direction of this axis is fixed in space—pointing always to the North Star.
 A vertical reference on the earth tilts at a rate of 360°/24hr times the cosine of its latitude. Thus:
15°/hr x cos(42.5) = 11.1 degrees per hour.