SAMPE Journal - January/February 2023 - 18

FEATURE / AUTOMATED EVALUATION METHOD
Figure 5 shows three-dimensional surface plots
of different regions in the target area containing a)
resin contamination and b) aged release coating.
The ML filter classified pixels in the plot into
three categories: resin contamination (yellow),
abhesive release coating (black), or aged release
coating (red). The plot's x- and y-axes correspond
to coordinates in the target area in units of pixels,
while the z-axis represents each pixel's RGB color
channel value. In addition to providing a visual of
the target zone, plots similar to Figure 5 may be
utilized in a production environment to quantify
the spatial coordinates, distribution, and percent
area of aged release coating or resin contamination
on a tool surface. During an inspection, these
values could then be used as go/no- go (i.e., pass/
fail) criteria.
Sparse Surface Free Energy Evaluation
Figure 6 shows side-view images of DI water
droplets immediately after solid-liquid contact
with a) abhesive release coating, b) aged release
coating, and c) resin contamination within the
tool's target area. The images in the figure were
captured using the low-fidelity portable digital
microscope. The top half of each image shows the
droplet resting on various tool surface conditions,
while the bottom half shows the droplet's reflection
in the Invar tool. When dispensed on the three
different surface conditions, DI water droplets
formed unique contact angles and distinct
spherical shapes. Droplets applied to abhesive
release coating exhibited the largest contact angles
and lowest width- to-height aspect ratios (i.e.,
spreading). Conversely, droplets dispensed on
resin contamination yielded the lowest contact
angles and highest aspect ratios. The contact angle
and shape a liquid droplet makes on a solid surface
directly results from intermolecular adhesive
interactions at the solid-liquid interface due to SFE
effects and internal cohesive forces in the liquid21
.
Since liquid properties of DI water remained
constant throughout the inspection procedure,
each droplet's unique contact angle and shape
must have been due to SFE effects. Therefore,
even without performing robust calculations, the
portable technique was able to identify distinct
qualitative differences in SFE between abhesive
release coating, aged release coating, and resin
contamination.
Figure 7 shows a plot of SFE versus DI water
contact angle and aspect ratio, with red points
representing calculated or measured values. A
response surface, shown in green, was fitted to
the data using a supervised ML method, Gaussian
Process Regression (GPR). GPR models possess
several benefits, one of which includes working well
on small datasets, such as the one represented in
Figure 7. The SFEs of abhesive release coating areas
were measured over an approximate range of 11-13
mJ/m2
. These exceptionally low values indicate the
majority of the target area and tool surface remained
highly abhesive after undergoing twenty processing
cycles without any RA reapplications. Aged release
coating and resin contamination SFEs were
measured over approximate ranges of 28-30 mJ/
m2
and 38-53 mJ/m2
, respectively. Surface patches
containing these higher SFE values were likely the
cause of decreased tool abhesion properties after
the twentieth autoclave cycle. Furthermore, the
largest SFE values measured on resin indicates that
contamination has more detrimental effects on
tool-part interaction and PIDs than aged release
coating. The standard deviations (SDs) of abhesive,
aged, and contaminated areas were approximately
0.4 mJ/m2
The notably larger
, 0.2 mJ/m2, and 3.5 mJ/m2
scatter in
, respectively.
contamination
measurements was likely due to a combination of
Figure 6. DI water droplets immediately after solid-liquid contact with a) abhesive release coating, b)
aged release coating, and c) resin contamination on a tool surface.
18 | SAMPE JOURNAL | JANUARY FEBRUARY 2023
www. sampe.org

http://www.sampe.org

SAMPE Journal - January/February 2023

Table of Contents for the Digital Edition of SAMPE Journal - January/February 2023