Cyclic testing has been carried out to characterize the fatigue and creep-fatigue behavior of Alloy 709, an advanced austenitic structural alloy. Specimens were machined from several plates of 709, subjected to different processing conditions and resulting in varying microstructures in the specimens, dependent on both the plate used for the machining and the location the specimen originated from within the plate. The cyclic testing has been conducted primarily at a single fatigue and creep-fatigue condition, which was at 650 °C and 1.0% total strain (+/-0.5% strain). For creep-fatigue, a 30 min. hold time at peak tensile strain was introduced. The combined results of the cyclic testing and post-test metallographic characterization suggest that the investigated microstructures, ultimately controlled by the processing conditions, have limited influence on the Alloy 709 fatigue cycle life. In contrast, some effect on the creep-fatigue cycle life due to the microstructure was observed, specifically that the presence of large grains may negatively impact the creep-fatigue resistance. The number of cycles to failure at the selected creep-fatigue condition varied by up to a factor of 5 for the specimens tested. Specimens with an optimum microstructure, a uniform distribution of grains with diameters less than approximately 200 ?m, resulted in a creep-fatigue resistance similar to that of 316-type stainless steel. The majority of the Alloy 709 cyclic test data has been previously reported in milestone reports - and the microstructural and processing details for the specific heats have been reported by Yamamoto et al. - This summary report includes creep-fatigue test results on three additional heats (011502F, 011502H, and 011539 (note, same plate as ORNL 011593)) for which the processing and microstructural details are available in previous reports.

The development of isochronous stress-strain curves is an important piece of the code case for Alloy 617. A 2-parameter creep equation was used to model creep based on test results at INL. The Ramberg-Osgood and Voce equations were used to generate hot tensile curves based on INL test results at different temperatures. Both the isochronous and hot tensile curves were then shifted to represent the average heat of Alloy 617.

Alloy 617 is the primary candidate material for the heat exchanger of a very high temperature gas cooled reactor intended to operate up to 950°C. While this alloy is currently qualified in the ASME Boiler and Pressure Vessel Code for non-nuclear construction, it is not currently allowed for use in nuclear designs. A draft Code Case to qualify Alloy 617 for nuclear pressure boundary applications was submitted in 1992, but was withdrawn prior to approval. Prior to withdrawal of the draft, comments were received indicating that there was insufficient knowledge of the creep and creep-fatigue behavior of Alloy 617 welds. In this report the results of recent experiments and analysis of the creep-rupture behavior of Alloy 617 welds prepared using the gas tungsten arc process with Alloy 617 filler wire. Low cycle fatigue and creep-fatigue properties of weldments are also discussed. The experiments cover a range of temperatures from 750 to 1000°C to support development of a new Code Case to qualify the material for elevated temperature nuclear design. Properties of the welded material are compared to results of extensive characterization of solution annealed plate base metal.

Isochronous stress-strain curves for Alloy 617 up to a temperature of 1000°C will be required to qualify the material for elevated temperature design in Section III, Division 1, Subsection NH of the ASME Boiler and Pressure Vessel Code. Several potential methods for developing these curves are reviewed in this report. It is shown that in general power-law creep is the rate controlling deformation mechanism for a wide range of alloy heats, test temperatures and stresses. Measurement of the strain rate sensitivity of Alloy 617 indicates that the material is highly strain rate sensitive in the tensile deformation range above about 750°C. This suggests that the concept of a hot tensile curve as a bounding case on the isochronous stress-strain diagrams is problematic. The impact of strain rate on the hot tensile curves is examined and it is concluded that incorporating such a curve is only meaningful if a single tensile strain rate (typically the ASTM standard rate of 0.5%/min) is arbitrarily defined. Current experimentally determined creep data are compared to isochronous stress-strain curves proposed previously by the German programs in the 1980s and by the 1990 draft ASME Code Case. Variability in how well the experimental data are represented by the proposed design curves that suggests further analysis is necessary prior to completing a new draft Code Case.

Creep of Alloy 617, a solid solution Ni-Cr-Mo alloy, was studied in the temperature range of 1023 K to 1273 K (750 °C to 1000 °C). Typical power-law creep behavior with a stress exponent of approximately 5 is observed at temperatures from 1073 K to 1273 K (800 °C to 1000 °C). Creep at 1023 K (750 °C), however, exhibits threshold stress behavior coinciding with the temperature at which a low volume fraction of ordered coherent γ′ precipitates forms. The threshold stress is determined experimentally to be around 70 MPa at 1023 K (750 °C) and is verified to be near zero at 1173 K (900 °C)—temperatures directly correlating to the formation and dissolution of γ′ precipitates, respectively. The γ′ precipitates provide an obstacle to continued dislocation motion and result in the presence of a threshold stress. TEM analysis of specimens crept at 1023 K (750 °C) to various strains, and modeling of stresses necessary for γ′ precipitate dislocation bypass, suggests that the climb of dislocations around the γ′ precipitates is the controlling factor for continued deformation at the end of primary creep and into the tertiary creep regime. As creep deformation proceeds at an applied stress of 121 MPa and the precipitates coarsen, the stress required for Orowan bowing is reached and this mechanism becomes active. At the minimum creep rate at an applied stress of 145 MPa, the finer precipitate size results in higher Orowan bowing stresses and the creep deformation is dominated by the climb of dislocations around the γ′ precipitates.

Tensile testing has been conducted on Alloy 617 bar stock in the temperature range of room temperature - 1000°C. Repeatability for replicate tests is excellent and temperature dependent properties are consistent with previous observations for this alloy. Comparability to historical data is significant since modern mill practice incorporates an additional refining step, electro-slag re-melting, that has only recently become standard practice. The results are compared to those of a reference plate that has been extensively characterized previously in this program. These tests provided data for an alternative heat and product form that is required for the ASME Boiler and Pressure Vessel Code qualification to allow this material to be used in nuclear pressure vessels. The results also extend the temperature range over which the alloy has been characterized compared to current allowable stresses in the ASME Code for non-nuclear pressure vessel design. The bar stock generally has higher strength and ductility than the reference plate. Statistical analysis has been performed on recent tensile data determined in this program grouped with CEA (Commissariat à l'énergie atomique et aux énergies alternatives – the French research organization) for contemporary plate material and Oak Ridge National Laboratory data from a number of years ago to determine if they are significantly different than data from the original draft code case data generated at Huntington Alloys many years ago. A best least-squares fit of a polynomial was used, although a piecewise function can provide a better fit for both the yield and tensile strength of the material as a function of temperature. Analysis of the yield strength data shows the 95% confidence bounds of the new data set overlaps that of the original data set over the entire temperature range, indicating no difference in the two data sets and very little change in the design curve. The 95% prediction bound for yield strength provides a consistent, meaningful lower bound for yield strength, and would be a good candidate for minimum yield strength at temperature. The 95% confidence bounds for the tensile strength of the new and old datasets do not overlap above about 625°C, indicating a difference between the data sets at higher temperatures. Including the additional newer data would cause minor changes in the average tensile strength design curve, resulting in a slightly more conservative curve. The 75% prediction bound for tensile strength provides a consistent, meaningful lower bound for average tensile strength, and would be

(8,000 – 20,000 hours) creep tests of Alloy 617 weld metal. While most of this testing is ongoing, the current state of the tests indicate that the multi-axial stress state imposed by the notch geometry does not negatively impact the creep rupture life of the Alloy 617 weld metal. While the weld metal is notch strengthening in short-term (1,000 – 2,000 hours) testing, it is not clear if this characteristic will continue to hold for intermediate and long-term testing. Ongoing tests will provide additional information to address this concern of a crossover from notch strengthening to notch weakening for intermediate and long-term creep lives.

Residual stresses were calculated from the curvature of coating-substrate coupons using three different models: a simple two-beam elastic model, the Tsui–Clyne progressive deposition model, and the Tsui–Clyne progressive deposition model with substrate plasticity. The coatings studied were metallic and prepared by high-velocity oxy-fuel (HVOF) thermal spraying. The calculated stresses were compared to those measured on the same coupons using X-ray diffraction (XRD) techniques. Coating surface stresses calculated using the two-beam elastic model disagreed with those measured using XRD for coupons with significant curvature. Trends in residual stresses (with varying coating and substrate thickness, substrate material, and HVOF spray particle velocity) predicted by the elastic and elastic-plastic versions of the Tsui–Clyne progressive deposition model agreed with the trends measured by XRD. The magnitudes of stresses calculated using the Tsui–Clyne model agreed with the XRD measurements for coatings sprayed at low particle velocities but were significantly more compressive for coatings sprayed at higher velocities. Accounting for substrate plasticity in the Tsui–Clyne model improved the agreement with the XRD results, but only slightly.

Finite element simulations are used to examine surface cracks at regions of local curvature (corners or convolutions) in protective oxide scales. Stresses are generated during cooling from oxide formation temperatures. Three different modeling approaches are employed, since each adds some insight to crack behavior. For the first, a series of standard static analyses with varying crack lengths is used to approximate crack motion. Next, a simple node-release technique is used, permitting dynamic crack growth along an assumed path. Finally, a model based on an arbitrary crack path is employed, wherein the crack path is included as an unknown and is part of the solution. To quantify geometric effects, three different ratios of corner radii to scale thickness are considered. Further, the influence of the substrate material is investigated by considering both perfectly-plastic and work-hardening behavior. The computed stress-intensity factor at the crack tip is compared to the fracture toughness of the scale material to predict crack growth. Simulations indicate that sharper corners and lower substrate yield strengths increase crack growth potential. Reductions in the stress-intensity factor with increasing crack length are observed that result from the constraining effects of the substrate. Predictions of crack trajectory indicate initial crack motion perpendicular to the free surface of the scale, followed by a near 90° turn, resulting in a crack path nearly parallel to the free surface.

The fracture behavior of a Ni-Al2O3 Functionally Graded Material (FGM) was investigated using Phase Shifted Moire Interferometry (PSMI). A relatively thick (similar to 8mm) tri-layer FGM compact tension specimen geometry was used. Experimental result provide quantitative displacement data for comparisons to numerical models of near crack tip deformations. Fracture path analysis verified that the crack. propagated through the brittle alumina phases or as intergranular fracture around the ductile nickel particles. These results demonstrate the value of PSMI to measure the displacement fields surrounding advancing cracks in graded structures.

Abstract Thermally sprayed coating characteristics and mechanical properties are in part a result of the residual stress developed during the fabrication process. The total stress state in a coating/substrate is comprised of the quench stress and the coefficient of thermal expansion (CTE) mismatch stress. The quench stress is developed when molten particles impact the substrate and rapidly cool and solidify. The CTE mismatch stress results from a large difference in the thermal expansion coefficients of the coating and substrate material. It comes into effect when the substrate/coating combination cools from the equilibrated deposit temperature to room temperature. This paper describes a laser-based technique for measuring the curvature of a coated substrate and the analysis required to determine residual stress from curvature measurements. Quench stresses were determined by heating the specimen back to the deposit temperature thus removing the CTE mismatch stress. By subtracting the quench stress from the total residual stress at room temperature, the CTE mismatch stress was estimated. Residual stress measurements for thick (>1mm) spinel coatings with a Ni-Al bond coat on 304 stainless steel substrates were made. It was determined that a significant portion of the residual stress results from the quenching stress of the bond coat and that the spinel coating produces a larger CTE mismatch stress than quench stress.

Alumina scales that grow during oxidation of FeCrAl alloys can develop a convoluted morphology. Although convolution relieves the overall growth stress, high thermal stresses develop locally and can be detrimental to the scale or interface integrity. Ruby fluorescence measurements and finite element simulations are used to examine residual thermal stresses and strains that result when the convoluted scales are cooled to room temperature. Unlike a flat scale that is in biaxial compression, a convoluted scale contains significant gradients, with tensile stress components along the outside and near the interface of the convoluted peaks. The experimental results are in good agreement with model calculations and provide much needed verification of the model assumptions. Because the ruby fluorescence technique provides only the hydrostatic stress averaged over an excited volume that includes the entire alumina scale thickness, modeling provides detail and insight to the experimental measurements.

Residual stress near edges and corners of thermally grown alumina scaleswere investigated. In this study, an edge is the intersection of twoorthogonal flat surfaces and a corner is the intersection of three suchsurfaces. Microfluorescence measurements, performed on alloys withcomposition Fe–28Al–5Cr (at.%, bal. Fe) oxidized at 900°C,showed a large (>50%) reduction in hydrostatic stress in the vicinity ofedges and corners. Surprisingly, significant stress reduction persists outto distances twenty to fifty times the scale thickness from theedge. Finite-element analysis calculations confirm the experimental resultsand provide a considerably more detailed picture of the stress distributionand its components and show that much of the observed stress reduction nearan edge, is due to plastic deformation of the underlying metal.

Finite element simulations are used to examine surface cracks at regions of local curvature (corners or convolutions) in protective oxide scales. A series of static analyses with varying crack lengths is used to approximate a propagating crack through the scale thickness. Stresses are generated during cooling from oxide formation temperatures. Two different ratios of corner radii to scale thickness are considered to quantify geometric effects. The influence of the substrate material is investigated by considering both perfectly-plastic and work-hardening behavior. The computed stress-intensity factor at the crack tip is compared to the fracture toughness of the scale material to predict crack growth. Simulations indicate that sharper corners and lower substrate yield strengths increase crack growth potential. Reductions in the stress-intensity factor with increasing crack length are observed that result from the constraining effects of the substrate. Additional simulations designed to verify key modeling assumptions are described.

The scales that grow from oxidation often develop a convoluted morphology or interface pores. High thermal stresses can develop locally and are potentially detrimental to the scale or interface integrity. Finite element simulations are used to examine residual thermal stresses and strains that result when these deviations from a flat interface have formed, and the resulting geometry is subsequently cooled to room temperature. A variety of geometries will be considered for alumina scales on a FeCrAl substrate.

Finite element simulations are used to examine residual thermal stresses and strains in protective Al2O3 Scales on Fe3Al specimens during cooling from oxide formation temperatures. Geometrically, the simulations focus on regions of local curvature, either due to corners or convolution of the substrate surface. The effects of the corner radius of curvature, convolution wavelength/amplitude ratio, and scale thickness, as well as the impact of aluminide plasticity, are considered. Localized plasticity is found to have a major influence on net deformation and the magnitude and location of maximum stress. Based on the calculations of residual stress and strain distributions prior to the development of any localized damage, key implications about the tendencies for damage are addressed.

Finite element simulations are used to examine residual thermal stresses and strains in corner regions of protective Al2O3 scales on Fe3Al specimens, both during cooling from oxide formation temperatures and during subsequent thermal cycling. The effects of a corner's radius of curvature and oxide thickness, as well as the impact of aluminide plasticity, are considered. Localized plasticity is found to have a major influence on net deformation and on the magnitude and location of maximum stress. As the ratio of corner curvature to oxide thickness (rs/t) is reduced, stresses within the oxide corner shift from highly compressive to tensile and the location of the maximum principal stress moves from the substrate to the oxide scale. Based on these stress distributions prior to the development of any flaws, key implications about the tendencies for damage are addressed. The stress evolution during cooling and thermal cycling is presented; these results demonstrate the effects of temperature-dependent material properties. For the material behavior assumed in this study, thermal cycling does not cause significant stress relaxation.

Finite element simulations are used to examine residual thermal stresses and strains in corner regions of protective Al2O3 scales on Fe3Al specimens, both during cooling from oxide formation temperatures and during subsequent thermal cycling. The effects of a corner's radius of curvature and oxide thickness, as well as the impact of aluminide plasticity, are considered. Localized plasticity is found to have a major influence on net deformation and on the magnitude and location of maximum stress. As the ratio of corner curvature to oxide thickness (r(s)/t) is reduced, stresses within the oxide corner shift from highly compressive to tensile and the location of the maximum principal stress moves from the substrate to the oxide scale. Based on these stress distributions prior to the development of any flaws, key implications about the tendencies for damage are addressed. The stress evolution during cooling and thermal cycling is presented; these results demonstrate the effects of temperature-dependent material properties. For the material behavior assumed in this study, thermal cycling does not cause significant stress relaxation. (C) 1997 Elsevier Science S.A.

Experiments coupled with finite element modeling (FEM) have been used to examine residual stresses that arise in thermal spray samples. Substrate characteristics such as thickness, peak surface temperature, thermal gradients, and sample holder constraints have been examined in addition to the more commonly studied coating/substrate property mismatch effect. Elastic-plastic FEM considers temperature dependent thermal and mechanical properties during heating to the maximum surface temperature and subsequent cooling to ambient temperature, to calculate sample deformation, temperature distributions, and plastic strains in addition to residual stresses. Experiments utilize thermocouples located along the substrate radius and thickness to determine thermal gradients, and a specially designed sample holder to provide temperature control. Measured residual deformation at room temperature is compared to FEM results.

Grain sizes were measured after various heat treatments in three Fe3Al alloys having similar composition that were fabricated using the techniques of ingot metallurgy (cast and wrought), hot extrusion of prealloyed powder, and hot isostatic pressing (HIP) of elemental powders. The ingot metallurgy (I/M) material exhibited normal grain growth behavior at temperatures above 750 °C, in agreement with previous observations. Both powder metallurgy (P/M) materials displayed unusual resistance to grain growth compared to the I/M alloy. In the case of the prealloyed P/M material, the initial (recrystallized) grain size was larger than the initial grain size of the I/M material, although little grain growth was observed for heat-treatment temperatures up to 1100 °C. At higher temperatures grain growth occurred at a rate comparable to that observed to the I/M alloy. The elemental powder P/M material exhibited similar grain growth behavior to the prealloyed P/M material, although the initial (as-HIPed) grain size was considerably smaller. Transmission electron microscopy (TEM) indicated that the grain growth resistance of the P/M materials could be attributed to grain boundary pinning by oxide particles presumed to originate from the powder particle surfaces. The difference in the stable grain size between the prealloyed and elemental powder P/M materials was attributed to the nature of the particle dispersions resulting from processing.