120 likes | 249 Views
SANS examination of precipitate microstructure in creep-exposed single-crystal Ni-base superalloy SC16. P. Strunz 1,2 , G. Schumacher 1 , W. Chen 3 , D. Mukherji 4 , R. Gilles 5 and A. Wiedenmann 1. 1 Hahn-Meitner-Institut, Glienickerstr. 100, 14109 Berlin, Germany
E N D
SANS examination of precipitate microstructure in creep-exposed single-crystal Ni-base superalloy SC16 P. Strunz1,2, G. Schumacher1, W. Chen3, D. Mukherji4, R. Gilles5 and A. Wiedenmann1 1Hahn-Meitner-Institut, Glienickerstr. 100, 14109 Berlin, Germany 2Nuclear Physics Institute, 25068 Řež near Prague, Czech Republic 3Bundesanstalt für Materialforschung und -prüfung, Unter den Eichen 87, 12205F Berlin, Germany 4Technische Universität Braunschweig, 38106 Braunschweig, Germany 5Technische Universität Darmstadt, Petersenstr. 23, 64287 Darmstadt, Germany
nickel base superalloys - rafting • High-temperature+ slow-strain-rate exposure: an important regime of operation of turbine blades made of Ni-base superalloys (precipitation hardened alloys: g’ precipitates in g matrix). • In this regime: rafting (the g’ morphological change which significantly influences the lifetime of the blades) • Rafting: the initial cuboidalg’ precipitates coarsen to a plate like or needle like morphology (the rafts) • Very complex phenomenon depending on the g/g’ lattice misfit, rate and temperature of deformation, initial microstructure, orientation ...
objectives • Rafting: simultaneous particle agglomeration and particle growth but the mechanisms of raft formation not fully understood at present • Small-angle neutron scattering (SANS) measurement of initial stages of the morphological changes in the bulk material: help to resolve some of the questions in the rafting phenomenon • The aim: to study the initial stagesof morphological changes during the formation of rafted g’-precipitate structure in the SC16 single crystal Ni-superalloy after high temperature creep
experimental • SC16 single crystal bars: deformed at 950°C to different strains (tensile stress of 150 MPa along [001] crystal direction, strain rates <10-6 s-1) • SEM, strain 0.1% • SEM, strain 0.5%
experimental • V4 facility of BENSC in HMI Berlin • sample-to-detector distance 16 m • l = 19.4 Å (“low-Q range”) and l = 6.0 Å (“large-Q range”). • “low-Q range”: low flux of source => measured without the beam-stop normally protecting 2D PSD against overloading • samples of thickness 1.5-2 mm for SANS were cut out of these bars after unloading and cooling to the room temperature • The normal direction to the samples was parallel to [010]
measured data (SC16, creep, low-Q) • Measured (gray scale) and fitted (solid lines) differential cross-sections dS/dW (in cm-1sr-1, logarithmic scale) • strains 0, 0.1, 0.5 and 1.4% • low-Q region: the effect dominated by the scattering from g' phase • w-scan: fitted at once (3 meas.)
measured data (SC16, creep, large-Q) • SANS pattern measured in large-Q range for the most deformed sample (1.4% strain) • streaks in <320> directions => presence of topologically close packed (TCP) phase • scattering from TCP is comparable with the scattering from g' in this Q-range
microstructural model and evaluation • Anisotropic SANS evaluation: direct 3D "binary map" modeling followed by Transformed Model Fitting • The used model: partially ordered cubiodal and/or plate-like particles • Realistic approximation of a partial ordering: a Monte Carlo based simulation of positions and sizes of particles • A long-range size distribution included into one 3D "binary map" • The model of the individual cuboidal particle: according to the model introduced by Schneider et al. (J. Appl. Cryst.33, 465-468 (2000)) • In 3D space, the point belongs to the particle when the following is fulfilled: • x0, y0, z0 ... coordinates of the center; Rx, Ry, Rz ... "radii” • b defines shape: sphere or ellipsoid for b=1; it becomes more cuboidal, rod-like or plate-like when b decreases towards zero (exact cube or block with rectangular edges for b -> 0)
models resulting from the fit • two models used to fit the SANS data: the cuboidal one (Rx=Ry=Rz) and the plate-like one (Rx=Ry>Rz) - rafts • Real-space models resulting from the SANS-data evaluation (corresponding to the presented fits) • For 0.5% strain, both models were necessary to apply simultaneously • The gray scale: a slice of the 3D model having the thickness approximately equal to twice mean distance between precipitates was projected to 2D assuming a certain transparency of the modeled precipitates 0.0 % 0.1 % 0.5 % 1.4 %
results and discussion • For the deformations 0.0% and 0.1%, cuboidal precipitates were sufficient to describe the observed SANS patterns • A combination of both cuboidal and plate-like precipitates was necessary to apply for the deformation 0.5% • The data from 1.4% deformation could be successfully described by plate-like rafts alone • ----------------------------------------------------------------------- • Nearly no indication of rafting after deformation to 0.1%. However, the change of the shape of precipitates during this initial deformation period occurred: originally rather cubic precipitates transform to cuboids at 0.1% strain • Indications that diffusion flow during initial stages of creep can cause such rounding were published earlier
results and discussion • The evolution of the refined shape parameter b for cuboids • The evolution of the proportion "individual cuboids - rafts"
conclusions • Presented SANS: bulk information on g'-phase morphology changes during creep deformation of SC16 • Evolution of precipitate microstructure: three stages • First stage: no rafting occurs but the precipitates become significantly more rounded • Second stage: the rafts develop as more and more cuboidal precipitates agglomerate with each other • Transition between 1st and 2nd stage: between 0.1 and 0.5% • Above 1.4% strain, practically all precipitates in the bulk of the sample are rafted ACKNOWLEDGEMENT Two of the authors (R. Gilles and D. Mukherji) thank BENSC for support enabling to carry out the SANS experiment.