EXECUTIVE SUMMARY

This report presents the results of an experimental program dealing with the rheology of fresh concrete. The three main goals were:

• Obtain rheological data on concretes of various mixture compositions;
  
• Establish models to link mixture composition with rheological parameters, and
  
• Develop a simple field test for determination of rheological parameters.

Approximately 80 mixtures (mortars and concretes, with and without HRWRA) were formulated and tested. The rheological characteristics of the mixtures were measured using a BTHREOM rheometer, with parallel tests using the slump cone. It appears that the relationship between the torque and rotational speed of the rheometer is not linear, implying a non-linear relationship between stress and strain rate for concrete. A description of the rheological behavior of the material that is better than the usual linear Bingham model is provided by the Herschel-Bulkley model in the form  = '₀  +a ^b, where is the shear stress, is the shear strain rate imposed on the sample, and '₀ , a and b are characteristic parameters of the concrete being tested. Among other advantages, the non-linear model avoids the problem of a negative yield stress, which is sometimes encountered when the Bingham model is fitted to the test data. The "plastic viscosity" is defined from the characteristic parameters. However, for a certain number of applications, the Bingham approach may be retained as a first approximation.

Using the data obtained, the rheological characteristics of the mixtures were expressed as a function of their composition. A physical interpretation of the results is proposed. According to this interpretation, the yield stress term is a manifestation of friction between solid particles, while the "plastic viscosity" term results from viscous dissipation due to the movement of water in the sheared material. General forms for the models were deduced and the models were calibrated using the experimental data. The plastic viscosity appears to be controlled essentially by the ratio of the solid volume to the packing density of the granular mixture (including the aggregates and the cement). The yield stress is the result of an accumulation of contributions of each granular class, these contributions involving the size and roughness of the particles and their affinity for HRWRA. In addition to their predictive power, which must be confirmed with other experimental data, the models make it possible to understand several general tendencies observed in the rheology of cement-based materials when mixture proportions are varied, such as the influence of w/c ratio, cement content.

Finally, the development of a simple field test for estimating yield stress and plastic viscosity is described. The test in measuring the are described. The results of this test for all mixtures used in the program were analyzed. Semi-empirical models are proposed for estimating yield stress and plastic viscosity based on the results of this modified slump test. The area of application of the test for evaluation of the plastic viscosity is limited to concretes with a slump of 120 to 260 mm. If the validity of the modified slump test is confirmed in the future, it could be used for everyday quality control of concrete in the field.