TSRST heterogeneous modelling for investigating failure depending on asphalt mix design and experimental conditions

Abstract

Thermal Stress Restrained Specimen Test (TSRST) is used to determine low temperature cracking susceptibility of asphalt mixture. In this test, a specimen is subjected to a constant temperature decrease rate. Due to the prohibited thermal shrinkage, cryogenic stress is built up in the sample. The progression of stress causes failure in the specimen. There is a particular experimental limitation to measuring time-dependent stress. This paper aims to overcome this limitation and develop a numerical approach to estimate the specimen's local stress and strain value. It consists of heterogeneous multi-scale modelling of asphalt mixture. The heterogeneous scales are mortar and hot mix asphalt. The asphalt mixture scale is simulated as a triphasic model containing viscoelastic matrix, elastic inclusions, and air voids. Granular skeletons are generated using a random object modeller software or MOA (Modeleur d’Objet Aléatoire). These models correspond to a specific aggregate grading curve. After that, the heterogeneous geometries are imported into ABAQUS finite element code. The coupled thermal and mechanical numerical simulations are realised at different temperatures. Results have shown that the heterogeneous numerical model developed in this study can precisely predict time-dependent stress at different temperatures. Localising the critical zone with maximum stress within the matrix is considered in this study. The intensity of local stress and strain values are determined to specify the zones which are susceptible to damage. It was observed that the local stress is almost four to five times greater than the average failure stress at the global scale.

GRAPHICAL ABSTRACT

KEYWORDS:

• TSRST
• heterogeneous approach
• viscoelastic
• finite element modelling
• local stress
• biphasic

Acknowledgment

The authors would like to express his gratitude towards Rilem TC PIM TG2 (low-temperature performance of mixture) for the experimental results used in the present work.

Disclosure statement

No potential conflict of interest was reported by the author(s).