Skip to main navigation Skip to search Skip to main content

A unified study of crack propagation in amorphous silica: Using experiments and simulations

  • C. L. Rountree
  • , S. Prades
  • , D. Bonamy
  • , E. Bouchaud
  • , R. Kalia
  • , C. Guillot

Research output: Contribution to journalArticlepeer-review

52 Scopus citations

Abstract

Atomistic aspects of dynamic fracture in amorphous silica are investigated with molecular dynamics (MD) simulations. Simulations on amorphous silica were performed for two system sizes, 15 million and 113 million atoms. Crack propagation in these systems is accompanied by nucleation and growth of nanometer scale cavities up to 20 nm ahead of the crack tip. Cavities coalesce and merge with the advancing crack to cause mechanical failure. This scenario was also observed experimentally during stress corrosion ultra-slow fracture of glass using atomic force microscopy (F. Célarié et al., Phys. Rev. Lett. 90 (2003) 075504; S. Prades, D. Bonamy, D. Dalmas, E. Bouchaud, C. Guillot, Int. J. Sol. Struct. 42 (2004) 637). This mechanism has macroscopic consequences in terms of sample life-time and deformation field. The morphology of the fracture surfaces has also been studied by calculating the height-height correlation function. In general experiments reveal two universal roughness exponents, 0.5 for small length scales and 0.8 for large length scales. The MD simulations of the 15 million and 113 million atoms system find the first roughness exponent (0.5), but the second exponent (0.8) occurs over length scales inaccessible to MD simulations. Finally, the 113 million atoms simulation was used to map out the morphology and dynamics of the whole crack front. © 2006.
Original languageEnglish
Pages (from-to)60-63
Number of pages4
JournalJournal of Alloys and Compounds
Volume434-435
Issue numberSPEC. ISS.
DOIs
StatePublished - May 31 2007
Externally publishedYes

Fingerprint

Dive into the research topics of 'A unified study of crack propagation in amorphous silica: Using experiments and simulations'. Together they form a unique fingerprint.

Cite this