Rate effects and load relaxation in static fracture of concrete

Zdenek P Bazant*, Ravindra Gettu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

169 Scopus citations


Reports an experimental study of the fracture of concrete at various crack mouth opening displacement (CMOD) rates with time to peak loads ranging from about 1 sec to 3 days (over five orders of magnitude). Tests were conducted on three-point bend specimens of three sizes in the ratio 1:2:4. Quasi-elastic fracture analysis, based on the effective modulus from creep theory, is used to evaluate the results according to the size effect method. The fracture toughness is found to decrease in agreement with the trend known for the dynamic range. The effective length of the fracture process zone is found to decrease with increasing rate, which implies increasing brittleness and a shift toward linear elastic fracture mechanics behavior for slow loading. Load relaxation at constant CMOD in the prepeak and post-peak stages of fracture tests was also investigated. The response tends to a straight line in the logarithm of elapsed time, and the post-peak relaxation is nearly twice as strong as the linear viscoelastic relaxation of unnotched specimens. The difference between these two relaxations must be caused by time-dependent processes in the fracture zone. The results reveal that in concrete there is a strong interaction between fracture and creep, which might cause the load-carrying capacity of structures with cracks to decrease significantly with load duration. However, extrapolations to loads beyond several days of duration would be speculative.

Original languageEnglish (US)
Pages (from-to)457-468
Number of pages12
JournalACI Materials Journal
Issue number5
StatePublished - Sep 1 1992

ASJC Scopus subject areas

  • Building and Construction
  • General Materials Science


Dive into the research topics of 'Rate effects and load relaxation in static fracture of concrete'. Together they form a unique fingerprint.

Cite this