JPT Nº22  Sponsors & Supporters

 

 

INTRODUCTION

 

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., 2017).

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.

Back                                                                                                                                                                                                                 Next