7.3.1 When the effect of heat transfer is being measured, it
is acceptable to cool the test specimen to simulate end use
conditions. However, it should be realized that heat transfer is
undoubtedly the most difficult variable to simulate or control.
To develop fatigue data under refrigeration or at isothermal
conditions is only of value if the material will be used under
similar conditions. Artificial cooling shall not be used to force
the material to fail mechanically when thermal failure is the
controlling mechanism in the application