Quantitative assessment of microstructural changes of hydrated cement blends due to leaching and carbonation, based on statistical analysis of image data

Orkun Furat*, Andre Baldermann, Claudia Baldermann, Martin Dietzel, Volker Schmidt

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

The evolution of the microstructure in corrosive environments plays a key role for the performance and durability of cementitious materials, however, appropriate methods to quantitatively describe microstructural alterations are limited. Here, statistical analysis of microscopic data is used to describe changes in porosity, continuous and individual pore size distributions of reactive cement phases during leaching and carbonation of hydrated cement blends. Therefore, BSE images obtained from (un)damaged areas of the cement blends were segmented using image processing techniques, followed by geometrical characterization and quantitative evaluation of the microstructural response(s). It is shown that the dissolution of portlandite generates a high meso- and macro-porosity (> 100-4000 nm pores), whereas precipitation of C-(A)-S-H and Ca-carbonate polymorphs leads to a densification of the microstructure, i.e., reducing the fine meso- and micro-porosity (< 500 nm pores). Cement blends made with hydraulically active SCMs and chemically poorly reactive carbonate fillers performed better than pure (OPC-based) cement paste.

Original languageEnglish
Article number124370
JournalConstruction and Building Materials
Volume302
DOIs
Publication statusPublished - 4 Oct 2021

Keywords

  • Blended cements
  • Leaching
  • Microstructure
  • Statistical image analysis
  • Supplementary cementitious materials

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • General Materials Science

Fingerprint

Dive into the research topics of 'Quantitative assessment of microstructural changes of hydrated cement blends due to leaching and carbonation, based on statistical analysis of image data'. Together they form a unique fingerprint.

Cite this