# Machine-Learning-Assisted Viscoelastic Characterization of PC/ABS Blends via Multi-Frequency Dynamic Mechanical Analysis

## TL;DR

This paper uses machine learning and dynamic mechanical analysis to study and predict the viscoelastic behavior of a PC/ABS plastic blend.

## Contribution

The study introduces a physics-informed machine learning model (NeuralWLF) that outperforms data-driven models in cross-frequency predictions.

## Key findings

- NeuralWLF achieved R² > 0.92 for all targets in cross-frequency generalization.
- MLP performed best in interpolation with R²¯=0.989, but its performance dropped significantly with larger validation gaps.
- A physics-data crossover threshold was identified at a gap/FWHM ratio of ≈2, beyond which NeuralWLF outperformed data-driven models.

## Abstract

This study combines multi-frequency dynamic mechanical analysis (DMA) with machine learning (ML) to characterize and predict the viscoelastic properties of a commercial polycarbonate/acrylonitrile–butadiene–styrene (PC/ABS) blend. DMA temperature sweeps at four frequencies (1–10 Hz) in single cantilever mode yielded a glass transition range of 115.8–123.2 °C (E″ peak), frequency sensitivity of 7.18 °C/decade, and an apparent activation energy of 335±85 kJ mol−1. Time–temperature superposition master curves were parameterized with a six-term Prony series (R2=0.998). Four data-driven models (RF, XGB, SVR, MLP) and a physics-informed NeuralWLF model were evaluated through a hierarchical validation framework. Temperature-blocked CV ranked MLP (R2¯=0.989) above RF (0.950) for interpolation; LOFO validation revealed that NeuralWLF achieved the best cross-frequency generalization (R2>0.92 for all targets) with interpretable WLF parameters (C1≈12.2, C2≈51.7 °C). A systematic block size sweep (5–30 °C) revealed a validation inflation effect in which MLP tanδR2 dropped from 0.986 to 0.592 as the gap-to-FWHM ratio increased from 0.5 to 3.1, establishing the gap/FWHM ratio as a quantitative validation stringency criterion. A physics–data crossover was identified at gap/FWHM ≈2: beyond this threshold, NeuralWLF outperformed all data-driven models in tanδ prediction by up to +0.300 in R2, while curriculum learning (freezing the WLF layer for 300 epochs) further improved the most stringent 30 °C validation from R2=0.660 to 0.731. The integrated framework demonstrates that honest evaluation of DMA–ML models requires validation gaps exceeding the characteristic feature width and introduces a quantifiable physics-data crossover criterion for selecting between data-driven and physics-informed architectures.