96
D. W. Eastman et al.
sensed with photodiode arrays and converted into strain [48]. This technique enables
measurements of normal and transverse strains and allows for the direct measure of
Young’s modulus and Poisson’s ratio.
The most prominent noncontact method for measuring strain is digital image
correlation (DIC), which has been under development since the 1980s [49]. In
DIC, markers or speckle patterns are applied to the sample surface and the region
of interest is divided into subsets. Each of these subsets is tracked as the sample
deforms, and its correlation to the subset in the original image is determined using
a grayscale intensity function. The correlation coefficients of subsets of consecutive
images are calculated and plotted versus their position. The maximum correlation
coefficient is then located that is representative of the optimal fit between the subsets
of the two images. From this correlation, displacements can be determined and both
local and global strains can be calculated.
Recent progress in developing high-resolution speckle patterns has led to the
ability to observe strains at very localized intragranular regions. Kammers et al. used
gold nanoparticles to create speckle patterns with features on the order of 30 nm.
By imaging these particles in an SEM, DIC with an unprecedented resolution of
4 nm/pixel was reported [50]. One important consideration in performing DIC with
high resolution SEM images is distortions in SEM images, which require proper
control and the use of high magnification, low accelerating voltages, large spot
sizes, long dwell times, and low working distances [50]. Stinville et al. developed
a novel technique using the γ’ particles after a heat treatment in René 88DT as the
speckle pattern [51]. The nanometer-sized features allowed for high resolution DIC
in a fatigue sample to observe crack nucleation. Montgomery et al. developed a
technique that uses multilayered Au, Ti, and Ag-sputtered coatings reconfigured in
an NaCl solution to form DIC speckle patterns, allowing for submicron resolution
[52]. This technique was demonstrated to be applicable for multiple classes of
materials, including metal alloys, epoxies, and composites. These high-resolution
methods along with others provide a way to improve the resolution of DIC
measurements, and further improvements to the technique could be made with
higher resolution imaging with reduction in noise.
The material that was investigated as part of CEIMM is the polycrystalline Nibase superalloy René 88DT. The material has a nominal composition (wt %) of:
56.46 Ni, 13 Co, 16 Cr, 4 Mo, 4 W, 2.1 Al, 3.7 Ti, 0.7 Nb, 0.03 C, and 0.015 B. This
alloy is processed through a powder metallurgy route and as such contains a high
volume fraction of annealing twins. It exhibits exceptional strength, even at very
high temperatures, making it ideal for its primary use in turbine engines [53].
René 88DT is an interesting candidate for multiscale modeling because of the
features that exist within the material at multiple length scales. At the subgrain,
or intragranular, scale, the two-phase microstructure gives rise to properties that
affect the higher order scales, especially in terms of the high temperature properties
related to dislocation activity [53]. At the polycrystalline scale, the large volume
fraction of twins adds complexity in terms of being able to model these finer features
as well as the various boundaries that exist. A description of a CPFEM model to
approach this multiscale problem is given in [54]. The model endeavors to represent
D. W. Eastman et al.
sensed with photodiode arrays and converted into strain [48]. This technique enables
measurements of normal and transverse strains and allows for the direct measure of
Young’s modulus and Poisson’s ratio.
The most prominent noncontact method for measuring strain is digital image
correlation (DIC), which has been under development since the 1980s [49]. In
DIC, markers or speckle patterns are applied to the sample surface and the region
of interest is divided into subsets. Each of these subsets is tracked as the sample
deforms, and its correlation to the subset in the original image is determined using
a grayscale intensity function. The correlation coefficients of subsets of consecutive
images are calculated and plotted versus their position. The maximum correlation
coefficient is then located that is representative of the optimal fit between the subsets
of the two images. From this correlation, displacements can be determined and both
local and global strains can be calculated.
Recent progress in developing high-resolution speckle patterns has led to the
ability to observe strains at very localized intragranular regions. Kammers et al. used
gold nanoparticles to create speckle patterns with features on the order of 30 nm.
By imaging these particles in an SEM, DIC with an unprecedented resolution of
4 nm/pixel was reported [50]. One important consideration in performing DIC with
high resolution SEM images is distortions in SEM images, which require proper
control and the use of high magnification, low accelerating voltages, large spot
sizes, long dwell times, and low working distances [50]. Stinville et al. developed
a novel technique using the γ’ particles after a heat treatment in René 88DT as the
speckle pattern [51]. The nanometer-sized features allowed for high resolution DIC
in a fatigue sample to observe crack nucleation. Montgomery et al. developed a
technique that uses multilayered Au, Ti, and Ag-sputtered coatings reconfigured in
an NaCl solution to form DIC speckle patterns, allowing for submicron resolution
[52]. This technique was demonstrated to be applicable for multiple classes of
materials, including metal alloys, epoxies, and composites. These high-resolution
methods along with others provide a way to improve the resolution of DIC
measurements, and further improvements to the technique could be made with
higher resolution imaging with reduction in noise.
The material that was investigated as part of CEIMM is the polycrystalline Nibase superalloy René 88DT. The material has a nominal composition (wt %) of:
56.46 Ni, 13 Co, 16 Cr, 4 Mo, 4 W, 2.1 Al, 3.7 Ti, 0.7 Nb, 0.03 C, and 0.015 B. This
alloy is processed through a powder metallurgy route and as such contains a high
volume fraction of annealing twins. It exhibits exceptional strength, even at very
high temperatures, making it ideal for its primary use in turbine engines [53].
René 88DT is an interesting candidate for multiscale modeling because of the
features that exist within the material at multiple length scales. At the subgrain,
or intragranular, scale, the two-phase microstructure gives rise to properties that
affect the higher order scales, especially in terms of the high temperature properties
related to dislocation activity [53]. At the polycrystalline scale, the large volume
fraction of twins adds complexity in terms of being able to model these finer features
as well as the various boundaries that exist. A description of a CPFEM model to
approach this multiscale problem is given in [54]. The model endeavors to represent