Metodical error in determination of materials durability under nonisothermal loading

A method to evaluate a methodical error in the experimental determination of the durability is suggested. This error has arisen from the errors of measurement and regulation of temperature and stresses. It is shown on the particular example that the temperature error dependence of the error sensitivity in determination of the durability is four times higher than the stress error dependence, which specifies the most stringent requirements to the accuracy of the temperature reproduction in experimental studies