Work

Uncertainty model of viscosity properties in polymer adhesives based on interval process

Year: 2025

Type: article

Abstract: • Polymer adhesives exhibit complex time-varying viscosity influenced by various factors and shear rates. • Traditional viscosity models with probability theory are not effective for adhesive viscosity uncertainty. • An interval process model is proposed to accurately describes viscosity uncertainty in realistic conditions. • The model's accuracy is demonstrated in analyzing viscosity uncertainty in different adhesive types. Polymer adhesives are widely applied in engineering, but their viscosity exhibits significant fluctuations and discontinuities under varying shear rates, making traditional probabilistic models ineffective for uncertainty quantification. This study presents a non-probabilistic viscosity uncertainty prediction framework that integrates the interval process model with the uncertainty characterization. Two adhesives with distinct viscosities - a hot-melt adhesive with strong fluctuation and a UV adhesive with weaker variation - were tested (40 batches, 20 for model building, 20 for validation). The methodology involves: (1) serializing and normalizing the viscosity data to extract steady-state and fluctuation features; (2) constructing an ellipsoidal convex model to characterize uncertainty bounds and autocorrelation functions; and (3) applying an Interval Process Model utilizing the Karhunen-Loève (K-L) expansion truncated at 95% cumulative variance for uncertainty prediction. The model achieved high accuracy, with Pearson correlation coefficients up to 0.996, RMSE =142.12, NRMSE =0.0446, and MAPE =6.92% for hot-melt adhesive, and similarly strong results for UV adhesive. The proposed framework provides a data-efficient, robust, and physics-consistent tool for viscosity uncertainty quantification, with direct applications in process control, formulation optimization, and quality assurance in adhesive manufacturing.

Source: Results in Engineering

DOI: 10.1016/j.rineng.2025.107239

Authors: Liang Cao, S.G. Gong, Y.R. Tao, S.Y. Duan, G.Q. Chen

Institutions: Hunan Institute of Engineering, Xiangtan University, Hebei University of Technology

Language: English

Topic: Probabilistic and Robust Engineering Design

Subfield: Statistics, Probability and Uncertainty

Field: Decision Sciences

Domain: Social Sciences

Open Access status: gold

Estimated APC paid: $1,940

Funders: National Natural Science Foundation of China, National Key Research and Development Program of China