Embry-Riddle Aeronautical University


This paper investigates the factors affecting dimensional precision of consumer-grade 3D printing, attempting to isolate and mitigate sources of error. The focus is on creating engineering prototypes of, tooling for, or finalized instances of mechanical devices. A specific fused deposition modeling printer – the Ultimaker 2 – is analyzed in terms of meeting precise physical dimensions, consistent shapes, and predictable surface finish. Extensive trial and error resulted in removal of several sources of bias, with square test articles exhibiting a lower-than-anticipated mean percentage error of -0.387% (SD = 0.559), a value comparable to other modern manufacturing techniques. A full factorial design of experiments analysis was executed, with print speed, outer shell thickness, and model size as factors. Both main effects and interaction effects were analyzed for trends. Despite statements in the public literature citing certain factors as having clear precision implications, the observed trends exhibited a complex interaction between factors, defying the easy characterization of factors such as speed as being contributors or detractors from printing precision. Three-way Analysis of Variance was used to analyze statistical significance of the trends, revealing that the factors here are likely not the primary sources of the remaining approximately 0.3% dimensional variation observed. Some possible influences of this variation are recommended for future study, including z-wobble, improper arc compensation, and filament diameter inconsistency.