Paper Number
PO39 

Session
Poster Session

Title
Automatic construction of rheological master curves

Presentation Date and Time
October 13, 2021 (Wednesday) 6:30

Track / Room
Poster Session / Ballroom 1-2-3-4

Authors (Click on name to view author profile)

  1. Lennon, Kyle R. (Massachusetts Institute of Technology, Department of Chemical Engineering)
  2. McKinley, Gareth H. (Massachusetts Institute of Technology, Mechanical Engineering)
  3. Swan, James W. (Massachusetts Institute of Technology, Department of Chemical Engineering)

Author and Affiliation Lines (in printed abstract book)
Kyle R. Lennon1, Gareth H. McKinley2 and James W. Swan1
1Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02142; 2Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA

Speaker / Presenter 
Lennon, Kyle R.

Keywords
computational methods; rheology methods

Text of Abstract
We construct master curves by applying Gaussian process regression to model experimental data sets and then maximum a posteriori estimation to shifting parameters that "best" bring the statistical models derived from these data sets into registry with one another. We apply this methodology to various rheological data sets that exhibit parametric self-similarity, including time-temperature superposition of polymers and time-cure superposition of gels. The statistical framework for data superposition provides uncertainty estimates for the shifting parameters with minimal added computational cost, and is robust to experimental noise. Finally, we employ this automated approach to make forward predictions of data at an unmeasured state. This method for master curve construction, and subsequent forward predictions, is data-driven and non-parametric, and may be applied in a computationally efficient manner to a wide variety of data superposition problems.