110
D. W. Eastman et al.
in a material property, ensuring the results of the model at a larger volume have
no side effects with respect to this property of interest. However, there is a lot to
be gained in understanding what factors microstructurally affect the convergence of
properties to verify that the models capture this behavior at multiple length scales.
Observing these trends experimentally and investigating the role of microstructure
across length scales are needed to further develop and benchmark well-informed
models, such as the crystal plasticity finite element method (CPFEM), that account
for the microstructure of materials at the polycrystalline scale [54, 94].
Considerable work has been done to investigate size-scale effects on sample
strength in pure metals [95–103], but work on size-scale effects in structural alloys
such as René 88DT is more rare. In the current study, samples were tested across
multiple length scales, ranging from one grain through the sample thickness up to
bulk material, and their yield strengths were determined in order to evaluate the
divergence from the bulk strength. It was expected that variability in the data would
increase with decreasing sample size as discrete microstructures were realized.
Such scatter in mechanical response has been reported in the literature [60] for
various experimental designs [2, 93]. Experimentally capturing the size effects
over many length scales can provide a systematic approach to defining a RVE,
particularly in materials that have not been previously modeled in this manner [99].
The data collected in these experiments provide a better understanding of the role
of microstructure and size-scale effects in René 88DT and other FCC materials.
Additionally, these methodical studies aid in the development and benchmarking of
models that capture mechanical behavior.
Three different sample geometries were manufactured from foils of René 88DT.
Each had a uniform gage length, but different gripping strategies were employed.
The largest dog bone-shaped samples, with thickness and width ranging from
200–500 μm, were manufactured using wire EDM. Intermediate sized samples,
with thicknesses and widths between 50 and 100 μm, were machined using the
femtosecond laser. The smallest samples, with thickness and width of 20 μm,
were manufactured using automated FIB machining. The volumes tested in the
laser and FIB-machined samples provide discrete representation of mesoscale
microstructures for René 88DT.
In the Hemker research lab at Johns Hopkins University, there are multiple
microtensile testing setups that utilize similar configurations. These load frames
have varying load capacities of up to 1200 N and consist of 4 main components: an
air bearing, a load cell, a linear actuator, and a mechanism for gripping the samples
during a test. The air bearing assures alignment and eliminates friction from the
pull bar as it slides during the test. In-line load cells record load data at a rate of
10 points/second, and the screw-driven Zaber linear actuator retracts at a constant
speed to achieve nominal strain rate of 10 −4 per second.
For the dogbone-shaped samples sized 200 μm and larger, a custom set of high
strength titanium grips were utilized. One grip was attached to a pull bar that slid
within the air bearing while the other grip was mounted on a stationary block. At
the beginning of the uniaxial tension test, the sample self-aligned as the shoulders of
D. W. Eastman et al.
in a material property, ensuring the results of the model at a larger volume have
no side effects with respect to this property of interest. However, there is a lot to
be gained in understanding what factors microstructurally affect the convergence of
properties to verify that the models capture this behavior at multiple length scales.
Observing these trends experimentally and investigating the role of microstructure
across length scales are needed to further develop and benchmark well-informed
models, such as the crystal plasticity finite element method (CPFEM), that account
for the microstructure of materials at the polycrystalline scale [54, 94].
Considerable work has been done to investigate size-scale effects on sample
strength in pure metals [95–103], but work on size-scale effects in structural alloys
such as René 88DT is more rare. In the current study, samples were tested across
multiple length scales, ranging from one grain through the sample thickness up to
bulk material, and their yield strengths were determined in order to evaluate the
divergence from the bulk strength. It was expected that variability in the data would
increase with decreasing sample size as discrete microstructures were realized.
Such scatter in mechanical response has been reported in the literature [60] for
various experimental designs [2, 93]. Experimentally capturing the size effects
over many length scales can provide a systematic approach to defining a RVE,
particularly in materials that have not been previously modeled in this manner [99].
The data collected in these experiments provide a better understanding of the role
of microstructure and size-scale effects in René 88DT and other FCC materials.
Additionally, these methodical studies aid in the development and benchmarking of
models that capture mechanical behavior.
Three different sample geometries were manufactured from foils of René 88DT.
Each had a uniform gage length, but different gripping strategies were employed.
The largest dog bone-shaped samples, with thickness and width ranging from
200–500 μm, were manufactured using wire EDM. Intermediate sized samples,
with thicknesses and widths between 50 and 100 μm, were machined using the
femtosecond laser. The smallest samples, with thickness and width of 20 μm,
were manufactured using automated FIB machining. The volumes tested in the
laser and FIB-machined samples provide discrete representation of mesoscale
microstructures for René 88DT.
In the Hemker research lab at Johns Hopkins University, there are multiple
microtensile testing setups that utilize similar configurations. These load frames
have varying load capacities of up to 1200 N and consist of 4 main components: an
air bearing, a load cell, a linear actuator, and a mechanism for gripping the samples
during a test. The air bearing assures alignment and eliminates friction from the
pull bar as it slides during the test. In-line load cells record load data at a rate of
10 points/second, and the screw-driven Zaber linear actuator retracts at a constant
speed to achieve nominal strain rate of 10 −4 per second.
For the dogbone-shaped samples sized 200 μm and larger, a custom set of high
strength titanium grips were utilized. One grip was attached to a pull bar that slid
within the air bearing while the other grip was mounted on a stationary block. At
the beginning of the uniaxial tension test, the sample self-aligned as the shoulders of