Measuring techniques for autogenous strain of cement paste

Volumetric measurement of autogenous strain is frequently performed by placing the fresh cement paste in a rubber membrane submerged in water. The volume change of the cement paste is measured by the amount of water displaced by the submerged sample. Volumetric and linear measurements of autogenous strain should in principle give identical results. However, the measuring results from the volumetric method are typically 3-5 times higher than the results from the linear technique. In this paper, water uptake from the buoyancy bath through the rubber membrane is identified as the principal artefact of the volumetric method. Water absorption is driven by a lowering of the water activity in the cement paste due to dissolved salts in the pore fluid and to self-desiccation. From the moment of casting, significant water uptake was registered in all experiments. This water uptake influenced the volumetric measurements by the same order of magnitude as the autogenous strain itself. By performing the measurements in a paraffin oil bath instead of a water bath, this artefact was eliminated. Furthermore, volumetric measurements performed in paraffin oil gave almost identical results as linear measurements performed on the same cement pastes.