SEM-EDS and microindentation-driven large-area high-resolution chemomechanical mapping and computational homogenization of cementitious materials

Janille M. Maragh, Steven D. Palkovic, Asheesh Shukla, Oral Büyüköztürk, Admir Masic

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

The need for the chemomechanical characterization of composite materials of unknown composition arises in many fields, including in the study of ancient and biological materials. The exact properties and distributions of their constituent phases are often poorly understood, which makes the study of their effective properties more challenging. In this work, a chemomechanical framework for the homogenization of such composites is presented and applied to a series of cement-based mortars. First, backscattered scanning electron microscopy (BS-SEM) and SEM energy dispersive X-ray spectroscopy (SEM-EDS) are used to chemically characterize large sample regions at high resolution. Next, k-means clustering is used with the quantitative SEM-EDS elemental data to obtain the spatial distributions of each sample's chemically distinct constituent phases. Microindentation is used to obtain the local elastic properties of the aggregate phase, and microindentation and compression testing are used to obtain the local elastic properties of the binding phase of each sample. The data is then combined to form high-resolution maps of mechanical properties, which are used as the initial configurations of finite element models that are then subjected to uniaxial compression and pure shear deformations to investigate the microstructure response. The results are used to estimate the effective Young's modulus and shear modulus of each sample. It is shown that when the hardened cement paste phase is homogenized a priori using compression testing, the framework yields Young's modulus estimates with errors of 3.0–27.7%. A parametric study of the role of the reduced stiffness at the cement paste-aggregate interfaces shows that these errors may be further reduced to 1.4–6.0%.

Original languageEnglish (US)
Article number102698
JournalMaterials Today Communications
Volume28
DOIs
StatePublished - Sep 2021
Externally publishedYes

Bibliographical note

Funding Information:
The authors would like to thank Professor Tal Cohen, Professor Elsa Olivetti, and Professor John Williams and for their guidance in the formulation and completion of the presented work. We are also grateful to Dr. Thibaut Divoux for his expertise in micro- and nanoindentation, which was crucial for data collection and interpretation, and to Stephen Rudolph for his practical assistance in sample preparation and compression testing. We would also like to thank Wagner Petrographic (Lindon, UT, USA) for producing the polished sections used in this study.

Publisher Copyright:
© 2021 Elsevier Ltd

Keywords

  • Data clustering
  • Energy dispersive X-ray spectroscopy
  • Finite element modeling
  • Phase mapping

Fingerprint

Dive into the research topics of 'SEM-EDS and microindentation-driven large-area high-resolution chemomechanical mapping and computational homogenization of cementitious materials'. Together they form a unique fingerprint.

Cite this