2022 journal article

Physics‐based deep neural network model to guide electrospinning polyurethane fibers

Journal of Applied Polymer Science.

co-author countries: United States of America 🇺🇸
author keywords: coatings; fibers; membranes; porous materials
Source: ORCID
Added: September 20, 2022

Abstract Electrospinning is an inexpensive room‐temperature method of producing nanofibers from polymer granules. While it is quite easy to produce electrospun nanofibers, it is very difficult to control properties of the resulting fibers without conducting laborious trial‐and‐error experiments. In this work, a mass‐spring‐damper (MSD) model was used to simulate formation of electrospun fibers for different combinations of electrospinning conditions such as voltage, needle‐to‐collector distance, and polymer concentration. The sparse data from the CPU‐intensive MSD model were then used in developing a deep neural network (DNN) model that could guide the electrospinning experiment toward producing fibers with a desired diameter. The accuracy of the MSD‐DNN hybrid model was examined via comparison with experimental data obtained by electrospinning polyurethane fibers.