Benchmarking Parts to Compare Entry-Level Desktop 3D Printers to Stratasys Systems

Authors R.I. Campbell, D.J. de Beer, E. Pei, and M. Martorelli give us their insights on a variety of entry-level 3D printers in the recently published ‘Benchmarking of Maternal Extrusion Entry-Level 3D Printers.’ While 3D printers were originally created as rapid prototyping systems decades ago, they have progressed far past that, and are now able to create complex, functional parts.

Not all systems are made alike though, and in this study the authors compared less expensive hardware to those higher in price with a reputation for offering high performance.

Geometric benchmark part used for this study.

Entry-level 3D printers (EL3DPs) are included in this benchmarking study, ‘quantifying the dimensional capability’ of more common extrusion based EL3DPs. The authors explain that over the years as 3D printers have been growing increasingly popular, users still did not—and do not—have a lot to go on in terms of comparisons and benchmarking, so they made their own benchmark parts.

“Some of the parts were designed with a specific characteristic in mind, such as to evaluate surface finish, speed, or the effect of orientation upon mechanical properties,” stated the researchers. “However, most of the parts were presented as generic test pieces that could be used to evaluate the overall capability of any AM system.”

Today, such parts have evolved, to include details like overhangs and more. The team points out that standards for benchmarking still have not been created, and that reaching industrywide agreement could be difficult.

“For this research, it was decided that selecting a benchmark part with very fine features would not be an effective method for evaluation, as they would be unsuitable for extrusion- based methods. This discounted some of the more complex parts that had been previously designed. In addition, the high complexity of some benchmark parts would make measurement of all the features very resource intensive and time consuming,” stated the researchers. “Therefore, for the evaluation of geometric accuracy, it was decided that a simplified version of the generic benchmark part used by Mahesh et al would suffice.”

Geometric benchmark parts, from top left to bottom right are as follows: BFB 3000 Touch, Cube Generation 1, Cube Generation 2, PowerWasp01, RepRap, Up!, Dimension SST, Mojo.

Six EL3DPs were used to create eight different benchmark parts—also printed on Stratasys Mojo and Dimension SST machines. All parts were extracted from the printers afterward, with support structures removed, and readings written down in a logbook so they could be compared to the dimensional values from the initial CD models, allowing the researchers to figure out deviations.

“It should be noted that some features failed to build in some of the machines,” stated the researchers.

Summary of results from dimensional analysis

Most notably, ‘many’ of the parts had circular features that did meet the requirements for cylindricity. Others had ‘whiskers,’ while yet others had warped bottoms.

Extreme warping seen on part from BFB machine

In the second part of the evaluation, the samples were judged for aesthetics, judged solely on how nice they were to look at, rather than functionality of any sort. The research team showed them to 51 students spanning their second- and third-undergraduate years from the Product and Furniture Design course at De Montfort University in the United Kingdom.

“They were first asked to visually inspect the parts without touching them to score the visual quality. Next, they were asked to observe and handle the parts and score their perception of tactile and overall quality of the test pieces. In both cases, the scoring was from most acceptable (score 1) to least acceptable (score 4). All the scores for each part were then averaged and the results are shown in Table 3 (lower scores showing a higher level of quality).”

The BFB 3000 scored the worst out of all in terms of error, but the Dimension SST performed the worst out all 3D printers for aesthetics, and the research teams’ attributes much of this to the ‘whisker’ effect. Overall though, the researchers noted that none of the other machines were close in comparison to the quality or performance of the Stratasys 3D printers.

“From this study, it was found that the accuracy available from the best performing EL3DPs is approaching the capabilities of higher-priced machines but not yet equivalent to it. The authors speculate that this could be potentially due to the lack of build chamber temperature control on the lower-cost machines and the quality of the extruder head. It can be argued, therefore, that if functional prototype parts are required, EL3DPs are not yet fit-for-purpose,” concluded the researchers.

“A limitation to this study was that the experiments did not undertake repeated trials on each system and therefore the findings do not provide insight into the aspect of process variability. Further research could investigate the use of different materials to examine its impact upon warping and the use of different filament diameters and the nozzle sizes. Future work on a larger scale should examine how other parameters such as the temperature of the material, the fill density and the build pattern could affect the build quality of parts produced from EL3DPs.”

New hardware is constantly being created for 3D printing purposes, but today many do have their own niche from bioprinters to color powder 3D printers, compact metal 3D printers, and more. What do you think of this news? Let us know your thoughts! Join the discussion of this and other 3D printing topics at 3DPrintBoard.com.