Effect of Data Augmentation and Membership Functions on ANFIS-Based Tensile Strength Prediction of FDM-Printed PETG Parts

Fused deposition modeling (FDM) enables the fabrication of complex polymeric components; however, the tensile performance of printed parts is strongly influenced by processing conditions and can be difficult to predict when only a limited number of experimental observations are available. This study investigates the use of linear interpolation-based data augmentation combined with an adaptive neuro-fuzzy inference system (ANFIS) for predicting the tensile strength of FDM-printed polyethylene terephthalate glycol-modified (PETG) specimens. Layer thickness, infill density, and nozzle temperature were investigated using a Taguchi L9 experimental design, with three specimens fabricated for each condition and their mean tensile strength values used as the reference dataset. The experimental data were augmented in MATLAB using linear interpolation to generate datasets containing 200, 400, and 600 samples. Gaussian ANFIS models with three membership function configurations (M222, M333, and M444) were subsequently developed using 60/40, 70/30, and 80/20 training/testing ratios. Model performance was assessed using R2, MSE, and MEP, and a multi-criteria performance index was employed to determine the optimum configuration. The results showed that increasing the number of membership functions generally improved prediction accuracy, whereas increasing the augmented dataset size did not necessarily result in continuous performance gains. The optimum configuration was obtained with 400 augmented samples, an 80/20 training/testing ratio, and the M444 structure, yielding testing R2, MSE, and MEP values of 0.998935, 0.053107, and 0.00548758, respectively. The findings demonstrate that combining interpolation-based data augmentation with ANFIS provides an effective approach for predicting the tensile strength of FDM-printed PETG components when experimental datasets are limited.

Authors

Institutions

Publication Details

Journal
Black Sea Journal of Engineering and Science
Published
2026-11-15
DOI
https://doi.org/10.34248/bsengineering.2023280
Primary Topic
Additive Manufacturing and 3D Printing Technologies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Effect of Data Augmentation and Membership Functions on ANFIS-Based Tensile Strength Prediction of FDM-Printed PETG Parts

Ömer Çerlek
Black Sea Journal of Engineering and Science
Additive Manufacturing and 3D Printing Technologies
article

Effect of Data Augmentation and Membership Functions on ANFIS-Based Tensile Strength Prediction of FDM-Printed PETG Parts

Ömer Çerlek
article en

Abstract

Fused deposition modeling (FDM) enables the fabrication of complex polymeric components; however, the tensile performance of printed parts is strongly influenced by processing conditions and can be difficult to predict when only a limited number of experimental observations are available. This study investigates the use of linear interpolation-based data augmentation combined with an adaptive neuro-fuzzy inference system (ANFIS) for predicting the tensile strength of FDM-printed polyethylene terephthalate glycol-modified (PETG) specimens. Layer thickness, infill density, and nozzle temperature were investigated using a Taguchi L9 experimental design, with three specimens fabricated for each condition and their mean tensile strength values used as the reference dataset. The experimental data were augmented in MATLAB using linear interpolation to generate datasets containing 200, 400, and 600 samples. Gaussian ANFIS models with three membership function configurations (M222, M333, and M444) were subsequently developed using 60/40, 70/30, and 80/20 training/testing ratios. Model performance was assessed using R2, MSE, and MEP, and a multi-criteria performance index was employed to determine the optimum configuration. The results showed that increasing the number of membership functions generally improved prediction accuracy, whereas increasing the augmented dataset size did not necessarily result in continuous performance gains. The optimum configuration was obtained with 400 augmented samples, an 80/20 training/testing ratio, and the M444 structure, yielding testing R2, MSE, and MEP values of 0.998935, 0.053107, and 0.00548758, respectively. The findings demonstrate that combining interpolation-based data augmentation with ANFIS provides an effective approach for predicting the tensile strength of FDM-printed PETG components when experimental datasets are limited.

Black Sea Journal of Engineering and ScienceVol. 9(6)
Sakarya Uygulamalı Bilimler Üniversitesi (TR)
Additive Manufacturing and 3D Printing Technologies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Effect of Data Augmentation and Membership Functions on ANFIS-Based Tensile Strength Prediction of FDM-Printed PETG Parts — Ömer Çerlek · Black Sea Journal of Engineering and Science (2026) | TGRS Research Map | TGRS