Temps zéro de référence pour les mesures de retrait

Title: Temps zéro de référence pour les mesures de retrait
Author(s): J.-L. Granju, M. Sarkis, M. Arnaud, G. Escadeillas
Paper category : journal
Serial title: Materials and Structures
Abreviated Serial title: Mater. Struct.
ISSN: 1359-5997
Publisher: RILEM Publications SARL
Volume: 37
Issue: 271
Issue date: 2004-08-01
Publication year: 2004
Pages: 449 - 455
Total Pages: 7
Nb references: 15
Language: French

Abstract: Length changes of cement-based materials are a main concern for the practitioners.
Indeed, they are very often the cause of undesired disorders such as cracks.
Thus, their measurement and especially the measurement of shrinkage strain take a major importance.
Shrinkage measurements look easy and simple when they are performed on the hardened material.
They need a very careful operating process on a specially designed test fitting when fresh material, from mixing until 24 hours, is involved.
Nevertheless, this paper shows that, in both cases, the experimental protocols of the standards lead to a systematic artefact.
Especially, the shrinkage measured on hardened material, after demolding, is never in the continuity of the shrinkage measured in the previous period on the fresh material.
The origin-times proposed here restore this continuity.
Although this study was carried out on “normal” mortars (in agreement with the European standard EN 196-1: Methods of testing cement: Determination of strength), its focus is not to be limited to the case of cements characterisation.
It aims to provide a more reliable quantification of the shrinkage of all the cement-based materials, concerts and mortars, “normal” or not.
It relies on literature and experiment.
It enlightens that the origin-times usually taken into account for shrinkage measurements are not convenient.
After analysis, a set of relevant origin-times is proposed.
For measurement on fresh cement-based material (within 0-24 hours), the best is to take as origin-time the time of the maximum swelling measured in no-wind condition.
Else, the time of maximum temperature is admitted as a good approximation.
For measurement on hardened material, on the contrary of the procedure recommended in the present standards (for instance NF 15-433 or ASTM C596-01 and C490-00a) and taken as usual practice, the origin-time must be delayed long enough after the demolding time to allow the test specimens to recover a quasi-equilibrium (thermal and hydral).
In the case of 4?4?16 cm3 test specimens, the recommended delay is 3 hours.

Online publication: 2004-06-27
Classification: Scientific Reports
Publication type : full_text
Public price (Euros): 0.00
doi: 10.1617/14011