Neural Network Prediction of Natural Rubber Properties

Collaboration between: UW-Madison and SOAN Laboratorios

It is important to conduct research that promotes the preservation of ecosystems and natural resources for future generations. And with government entities beginning to implement a carbon tax, it is crucial for industry to embrace research for sustainable material alternatives. The Natural Rubber tree, known as the Heveas Brasiliensis tree resembles a sustainable production plant where it absorbs 3 kg of CO2, 1 L of water, and expels 3 kg of clean oxygen and 1 kg of latex (Figure 1) which can be coagulated, dried, and vulcanized into a diverse number of products.

Figure 1: Graphical depiction of how sustainable Natural Rubber production is

The Polymer Engineering Center focuses on conducting research to further the understanding on Natural Rubber and how additives influence viscoelastic properties, thermal properties and the processability of this sustainable material via characterization techniques and machine learning (Figure 2).

Figure 2. Graphical depiction on how machine learning will be utilized to predict viscoelastic and thermal properties of natural rubber blends based on the blend formulation.

The following research topics are being addressed:

  • Rheological Characterization
  • Viscoelastic Characterization
  • Thermo-mechanical Properties
  • Mechanical Devulcanization Study for Recyclability
  • Material Property Prediction via Machine Learning