Additive manufacturing (AM),
including desktop 3D printing, is a relatively new technology that allows for
the physical recreation of three-dimen-sional digital objects. Improvements in
AM have reduced the significant overhead costs and com-plexities associated with
the technology and expanded the user base beyond industrial pur-poses. Today,
even non-professional hobbyists and enthusiasts have been able to use AM
pro-cesses to create three-dimensional (3D) objects with relatively low costs
and with little training. This ease of use has also created opportunities for
professionals to utilize 3D printing to solve unique problems. In the field of
paleontology, 3D digitizing and printing have already been used to provide a
means of reconstruction, analyses, as well as unique and engaging forms of
outreach and education (e.g. Mallison, 2011; Rahman et al., 2012; Hasiuk, 2014;
Lautenschlager and Rücklin, 2014; Lautenschlager, 2016; Benoit et al., 2016;
Thomas et al., 2016).
Coupled with affordable,
high-quality digitization methods, 3D digitization and replication have strong
potential to revolutionize many of the ob-stacles in paleontology research
methods such as digital preservation, widespread dissemination of digitized
specimens (Tschopp and Dzemski, 2012), and challenges of working with large and/or
fragile specimens (Schilling et al., 2013; Mitsopoulou et al., 2015; Das et al.,
A variety of systems exist for the
creation of dig-ital representations of fossils, each with varying degrees of
affordability, resolution, ease of use, and speed. Many of these methods are
described in detail in Sutton et al. (2014). Surface-based techniques such as
photogrammetry and laser-texture scanning have the capacity to create highly
detailed digital representations of speci-men exteriors without damaging the
original material (e.g. Breithaupt and Matthews, 2001; Antcliffe and Brasier,
2008, 2011). Furthermore, in the case of computed tomography (CT) scan-ning,
digital restoration can be completed even if the specimen is still embedded in
matrix, or even inside of plaster jackets (e.g. Schilling et al., 2013).
With digital fossil
reconstructions, collaboration and sharing of data between researchers can also
be achieved electronically. Additionally, in cases where a physical, to-scale
replica of the specimen is ideal, creation of a tangible 3D model from the
digital file is considerably faster, easier, and potentially far less expensive
than re-questing a high-quality cast of the original.
Two commonly used techniques for
fossil digiti-zation are laser-texture scanning and structured-light scanning.
While each technique has varia-tions, smaller consumer-level laser-texture
scanners that are commonly utilized in research laboratories and museum
collections use a trian-gulation-based method. In such cases, a red laser is
fired at a targeted object and its position on the object is recorded by a
camera, triangu-lating the object's or point’s location in space (Sutton et al.,
2014; Figure 1A). Depending upon the desired resolution, settings can be
adjusted to modify the speed of the scanning laser, yield-ing more or fewer
captured points per in2. Projected structured-light scanning utilizes a
projected black-and-white pattern beamed onto a targeted object. The
displacement of the pro-jected pattern is then captured by a pair of cameras
aimed at the object, thus deriving a 3D point in space (Figure 1B). Depending on
desired resolution, settings can be adjusted to modify the shutter speed of the
cameras, yielding a faster or slower exposure time (milliseconds).
Additionally, 3D printed objects
are typically more durable than fragile cast replicas of original fossil
specimens (Tschopp et al., 2013). Analyses of anomalies in Egyptian mummies,
replication of objects for science outreach and communication, the
reconstruction of fossils from damaged field jackets, and investigations of
water flow through the structure of a digitally-scaled blastoid are all
situations in which 3D printing has provided unique perspectives of problematic
or delicate physical specimens (e.g. Rahman et al., 2012; Schilling et al.,
2013; Huynh et al., 2015; McKnight et al., 2015).
However, there are two major
obstacles to face when adopting 3D printing as a research tool: in-vestment
price and printer resolution. For researchers with a limited budget, the initial
in-vestment in digitization and 3D printing may present an infeasible hurdle.
While improve-ments and refinements in the technology have significantly reduced
the costs associated with 3D printing, the initial investment price may still be
out of reach for some researchers or smaller institutions.
A wide variety of 3D printing
units exist in the consumer and industrial markets. High-end in-dustrial units,
such as selective laser-sintering (SLS) and stereolithographic (SLA)
technologies use heat to fuse a thermoplastic powder (SLS) or liquid (SLA) to
structure tangible models. These units produce very high-resolution products
with build layers as thin as 20 μm and virtually no re-sidual print artifacts.
However, such units commonly cost up to tens of thousands of dollars (USD;
Gibson et al., 2015). Alternatively, lower-end consumer models commonly use
fused dep-osition modeling (FDM) technologies, where a thermoplastic filament
composed of acrylonitrile butadiene styrene (ABS) or polyacetic acid (PLA) is
extruded through a heated nozzle to build models of varying layer thicknesses,
commonly between 100-400 μm. This frequently results in the production of print
artifacts, such as ‘ribbed’ or ‘bumpy’ exteriors that may obfuscate minute
features and require post-printing finishing. These units usually cost a
fraction of the price of industrial units, making them attractive and
ac-cessible to many research laboratories, classrooms, and institutions. However,
they typ-ically lack the microscopic resolution of their industrial counterparts
(Gibson et al., 2010).
Figure 1: Utilized techniques of
digitization. A) Principles of laser-texture scanning, where a laser is
projected at an object, and the reflection is collected by a sensor; B)
Structured-light scanning, where a light pattern is projected on an object, and
a series of cameras triangulate the distance between points on the object.
Despite the differences in
technologies, abilities, resolution, and costs of digitization and additive
manufacturing systems, the level of accuracy needed is strongly dependent upon
the research or educational needs of the users. For example, phylogenetic coding
may require digitized or printed fossil reproductions of higher resolution than
reproductions needed for classroom instruc-tion or museum exhibition.
However, the fidelity maintained
from digitiza-tion to 3D printing can be measured and com-pared to assess at
which stages data may be lost, resulting in varying degrees of confidence in
dig-ital fossil reproductions. This study examines the fidelity of
paleontological data created using common digitization techniques and commercial
3D printer systems. Digital models created by both projected structured-light
scanning and tri-angulated laser-texture scanning were compared for deviation
between specimens to determine differences in fidelity between both scanning
methods. Additionally, these specimens were printed on different low-cost 3D
printers and subsequently measured to determine differences in fidelity based on
printer model and different printer settings. These measurements determine where
and how much data is lost in both the dig-itization and reproduction processes.
While the necessary level of
reproduction detail is dependent on the scope of a study or exhibi-tion needs,
the results of these experiments suggest that entry-level and commercial-grade
digitization and 3D printing units are useful for many paleontological research
and educational outreach needs.