Editorial Type:
Article Category: Research Article
 | 
Online Publication Date: 23 Jun 2022

Data-Driven Multiscale Science for Tread Compounding

,
,
,
,
,
,
, and
Page Range: 114 – 131
DOI: 10.2346/tire.22.21003
Save
Download PDF

ABSTRACT

Tread compounding has always been faced with the simultaneous optimization of multiple performance properties, most of which have tradeoffs between the properties. The search for overcoming these conflicting tradeoffs have led many companies in the tire industry to discover and develop material physics-based platforms. This report describes some of our efforts to quantify compound structures and properties at multiple scales, and their subsequent application in compound design. Integration of experiment and simulation has been found to be critical to highlighting the levers in data-driven multiscale compound tread design.

Keywords: rubber compounds; material modeling; material design; viscoelasticity; structure-property relationships; filler morphology; far-field rubber; filler-rubber interphase; scanning electron microscopy; structure correlation functions; finite element analysis; representative volume elements; statistical volume elements; dynamic moduli; materials informatics
FIG. 1 —
FIG. 1 —

A virtual design environment that connects the Material World to the Tire Design World through compound viscoelasticity.

FIG. 2 —
FIG. 2 —

Integral rubber concept, coupled to viscoelastic temperature sweeps: (a) Nordsiek's performance windows for tread performance. Performance arrows are displayed. (b) variable importance map of rolling resistance onto viscoelastic window; (c) variable importance map of wet traction onto a lower temperature window. Performance is determined by temperature/frequency windows in tan δ spectra. (a) is reproduced with permission from [4]. Copyright (1985) Huthig GmbH.

FIG. 3 —
FIG. 3 —

A general framework for determining filler structure and reconstruction. From experimental microscopy, a set of correlation functions are generated, and a statistically equivalent filler structure is built for a periodic compound cell.

FIG. 4 —
FIG. 4 —

Definition of different types of correlation functions.

FIG. 5 —
FIG. 5 —

Demonstration of the microstructure reconstruction process. Adapted with permission from [30]. Copyright (2013) Elsevier BV.

FIG. 6 —
FIG. 6 —

The basic framework of data-driven modeling methods (“target” refers to the correlations of the digitized medium; “actual” are the correlations of reconstructed material). Adapted with permission from [2]. Copyright (2012) Elsevier BV.

FIG. 7 —
FIG. 7 —

Interphase mechanical properties represented by master curve shifts to either stiffen or soften the mechanical response. DMA plot on the left taken from [2]. Copyright (2012) Elsevier BV.

FIG. 8 —
FIG. 8 —

(a), (b) Comparison of predicted tan δ curves with experimental data: here, the interphase properties are fixed while the interphase thickness is varied. Bottom graphic: visualization of the definition of the polymer-filler interaction, which we term the mechanical interphase, for FEA simulation. The thickness is in units of the geometry mesh, where the edge of one pixel is approximately 25 nm. Adapted with permission from [2]. Copyright (2012) Elsevier BV.

FIG. 9 —
FIG. 9 —

Comparison of predicted tan δ curves against experimental data: here the interphase thickness is fixed while the interphase properties, decades of shift (Ds), varied. Increasing Ds indicates increasingly longer relaxation times for the interphase. Reproduced with permission from [2]. Copyright (2012) Elsevier BV.

FIG. 10 —
FIG. 10 —

Finite element models (60 × 60 × 60) with representative microstructures and interphase gradients match experimental results at high filler concentrations, even when tan δ appears to shift to higher frequencies, or lower temperatures [2]. We observe that the peak position, at this high filler loading, is nearly the same with or without interphase, therefore controlled by the presence of the stiff phase. Adapted with permission from [37]. Copyright (2016) Elsevier BV.

FIG. 11 —
FIG. 11 —

Full evolution of a multiscale compound design environment, complete with experimental design concepts, on a multi-objective target space. This framework can be used for both analysis and compound discovery.

FIG. 12 —
FIG. 12 —

Elements of a complete materials informatics environment for compound design. Such environments merge knowledge from many technical disciplines, tying the elements into a design framework at its core.

Contributor Notes

1 Global Materials Science, and Tire and Performance Science Divisions, The Goodyear Tire and Rubber Company, Akron, Ohio 44306, USA and L-7750, Colmar-Berg, Luxembourg

2 Corresponding author

3 Department of Mechanical Engineering, The University of Connecticut, Storrs, Connecticut 06269, USA

4  Department of Materials Science and Engineering, Duke University, Durham, North Carolina 27708, USA

5 Department of Mechanical Engineering, Northwestern University, Evanston, Illinois 60208, USA

  • Download PDF