Abstract: Nine orthogonal experiments consisting of temperature, printing speed, layer thickness
were designed to investigate the factors and rules influencing fused deposition modeling (FDM)
precision. Surface roughness was measured using portable surface roughness apparatus. The
significant factor influencing surface roughness was determined by calculating the signal-to-noise
ratio and conducting the variance analysis. By means of the Taguchi method, the multiple regression
equation and exponential equation were employed to predict the surface roughness and to determine
the parameter combination of minimum surface roughness of FDM printouts. Results of the analysis
reveal the following factors in order of the extent of their influence on surface roughness: layer
thickness > temperature > printing speed. Validation experiments were undertaken regarding different
printing models and FDM printers to verify the validity and adaptability of the analysis results.
Experiment results show that the prediction result of the multiple regression equation outperformed
that of the exponential equation and Taguchi method. Furthermore, the above conclusions exhibit
broad adaptability to different printing models and FDM printers.

Key words: fused deposition modeling technology, surface roughness, significance analysis, prediction model