The results of conventional (Coffin-type) thermal-fatigue tests of Hastelloy N are reported. The plastic-strains induced by thermal stresses ranged from a hundred micro-units to more than ten-thousand microunits and correlated well with fatigue life. The slope of the plastic-strain fatigue relationship differs from the conventional slope of minus one half being about −0.9. The thermal-fatigue data are in good agreement with the isothermal (1300 and 1500 deg F) strain fatigue data available on this alloy. The same plastic-strain criterion for failure describes the results of tests where plastic flow is produced by yielding and a combination of yielding and stress relaxation. The plastic-strain energy per cycle versus life suggests that a constant plastic work to failure exists for this alloy. The implications of such a criterion are discussed.

An analysis of the conventional (Coffin-type) thermal-fatigue test is presented by describing an analytical model from which three methods for determining the plastic-strain range are derived. One of the methods for calculating the plastic strain does not require the knowledge or use of the thermal coefficient, modulus of elasticity, or stress range. The accuracy of this method is not affected by the variation of the thermal coefficient, modulus of elasticity, or yield strength during the heating and cooling periods. Plastic strains resulting from yielding and relaxation may be accounted for by this method. The errors of conventional methods are discussed. A method for plotting the stress-strain hysteresis loop is also given. From an analysis of the load-temperature loop, it is clearly shown how all of the significant variables can be determined without plotting a stress-strain hysteresis loop. The plastic-strain range, the mechanical-strain range (elastic plus plastic), the stress range, and the plastic-strain energy per cycle can easily be determined from the load-temperature loop. An alternate test is suggested that minimizes all of the major difficulties of the conventional test. An improved test specimen for the conventional test is also recommended.

The study of fatigue at elevated temperature requires the evaluation of material response to several load-time profiles in specified thermal-chemical-nuclear-electric fields. This review of test methods emphasizes the relationships between “pure low-cycle fatigue” and tensile, fracture toughness, creep, cyclic-creep, creep fatigue (interspersion), fatigue-cycle-with-hold, dynamic creep, fatigue crack growth, and high-cycle fatigue testing. The thermal and environmental control is discussed. Special attention is given to isothermal-uniaxial, bending, torsion, and combined stress tests. Cyclic thermal testing is reviewed. Strain measurement, signal conditioning, computer utilization, fatigue crack growth, vacuum and environmental control, nondestructive testing, acoustic emission, and fractography are reviewed as they pertain to fatigue at elevated temperature. Specimen design, gripping, and testing machine characteristics are also discussed.

No abstract

Low-cycle fatigue behavior of Hastelloy X at elevated temperature is presented. We studied the effects of test temperature, cycle frequency, and stress temperature arrangement with special attention given to the minimum ductility (short-time tension test) at or near 1300 F. Isothermal fatigue data at 800, 1300, 1500, and 1800 F are presented along with strain controlled thermal fatigue data for temperature ranges of 600 to 1800, 800 to 1300, and 1800 to 1300 F. The method of universal slopes can be used to describe the fatigue behavior under elevated isothermal and cyclic thermal fatigue conditions. The thermal fatigue behavior determined from controlled strain tests can be applied to the thermal cycle behavior of bundles of thin tubes joined to a rigid plate.