102
D. W. Eastman et al.
conductive in order to be able to pass a current to melt the material. As with the
FIB, there is a limit to the scale of microtensile samples that can be machined
using wire EDM. It becomes difficult to cut samples below a certain size because of
two main factors: the tension of the wire on the sample during the cutting process
and the surface roughness. In order for the wire EDM to cut, contact of the wire
with the sample must be maintained during machining, and the wire must be kept
tight in tension and a flow of water must also be maintained. For smaller, thinner
samples or more delicate materials, this can lead to the sample being bent during
the EDM process. One means of mitigating this is fixing the sample to a rigid
substrate using a conductive epoxy. Doing so protects the sample from bending
while still maintaining conductivity. However, there can still be challenges in using
this technique, as air gaps in the epoxy or between the epoxy and sample can cause
a short that will prevent the wire from cutting. Using a thicker, more rigid substrate
also requires more power in order to cut through both materials, which will limit
the quality of the actual sample. In addition, if the sample has to be sandwiched
between two plates to fully protect it, it can be difficult to align the sample prior to
machining.
Figure 3 presents side surface profiles captured using laser confocal microscopy
of three samples. Two of these samples were prepared using EDM, one with
machine settings and one with optimized settings, and the third is a surface
profile from a femtosecond laser-machined sample for reference. Using optimized
parameters and employing a multiple pass methodology for machining microtensile
samples reduces the average surface roughness from 22 to 2 μm [65, 67–70]. By
comparison, the surface roughness of a femtosecond laser-machined sample was
less than 1 μm.
The previously demonstrated surface roughness also poses a challenge when
machining microtensile samples as this roughness will have a much greater effect
at decreased sample sizes, as will be demonstrated in a later section. The roughness
can cause artifacts in the experimental data due to premature failure at surface flaws
or the creation of local stress concentrations. For most materials, the limit of the
sample size that can be created using wire EDM is only a few hundred microns,
if the roughness can be controlled and bending avoided. In the current study, the
smallest dogbone microtensile samples that were machined had a gage width and
thickness of 200 μm.
3.3 Femtosecond Laser Machining
Where the FIB is a valuable tool for machining microtensile samples with dimensions of less than 50 μm, and wire EDM is useful in machining microtensile samples
above a few hundred microns to the macroscale, there exists an intermediate length
scale where neither technique is effective. One tool that has shown potential for
machining samples at this intermediate length scale, however, is the femtosecond
laser. The use of femtosecond lasers for various material-removal processes has
D. W. Eastman et al.
conductive in order to be able to pass a current to melt the material. As with the
FIB, there is a limit to the scale of microtensile samples that can be machined
using wire EDM. It becomes difficult to cut samples below a certain size because of
two main factors: the tension of the wire on the sample during the cutting process
and the surface roughness. In order for the wire EDM to cut, contact of the wire
with the sample must be maintained during machining, and the wire must be kept
tight in tension and a flow of water must also be maintained. For smaller, thinner
samples or more delicate materials, this can lead to the sample being bent during
the EDM process. One means of mitigating this is fixing the sample to a rigid
substrate using a conductive epoxy. Doing so protects the sample from bending
while still maintaining conductivity. However, there can still be challenges in using
this technique, as air gaps in the epoxy or between the epoxy and sample can cause
a short that will prevent the wire from cutting. Using a thicker, more rigid substrate
also requires more power in order to cut through both materials, which will limit
the quality of the actual sample. In addition, if the sample has to be sandwiched
between two plates to fully protect it, it can be difficult to align the sample prior to
machining.
Figure 3 presents side surface profiles captured using laser confocal microscopy
of three samples. Two of these samples were prepared using EDM, one with
machine settings and one with optimized settings, and the third is a surface
profile from a femtosecond laser-machined sample for reference. Using optimized
parameters and employing a multiple pass methodology for machining microtensile
samples reduces the average surface roughness from 22 to 2 μm [65, 67–70]. By
comparison, the surface roughness of a femtosecond laser-machined sample was
less than 1 μm.
The previously demonstrated surface roughness also poses a challenge when
machining microtensile samples as this roughness will have a much greater effect
at decreased sample sizes, as will be demonstrated in a later section. The roughness
can cause artifacts in the experimental data due to premature failure at surface flaws
or the creation of local stress concentrations. For most materials, the limit of the
sample size that can be created using wire EDM is only a few hundred microns,
if the roughness can be controlled and bending avoided. In the current study, the
smallest dogbone microtensile samples that were machined had a gage width and
thickness of 200 μm.
3.3 Femtosecond Laser Machining
Where the FIB is a valuable tool for machining microtensile samples with dimensions of less than 50 μm, and wire EDM is useful in machining microtensile samples
above a few hundred microns to the macroscale, there exists an intermediate length
scale where neither technique is effective. One tool that has shown potential for
machining samples at this intermediate length scale, however, is the femtosecond
laser. The use of femtosecond lasers for various material-removal processes has