Next: Conclusion Up: Main Previous: Main
We first examined the flow of a suspension of mono-disperse spherical aggregates between solid cylinders under unidirectional vertical body force, representing gravity. This scenario is a model for the flow of SCC between rebars under its own weight. The diameter of the coarse aggregates or spherical particles is either one fifth or one-half of the distance between the rebars (Figure 1). These two cases obviously represent the extremes and not the likely gradation of coarse aggregates, but it is a good way to test the prediction: the concrete with the large aggregate (diameter 1/2 of the distance between the rebars) had a tendency to jam or stop flowing while systems with small aggregates (diameter 1/5 of the distance between the rebars) would flow freely without stopping. Further studies will consider the simulation of concretes with a broader distribution of aggregate size and shapes that is more representative of a real SCC.
Another application of this model for the concrete industry is the simulation of flow in various concrete rheometers. These rheometers, as shown elsewhere [9, 10] have different geometries and flow patterns. It was also shown in ref. 6 that the rheological properties measured by different concrete rheometers cannot be compared directly. The reason for these discrepancies is not clear at this time, although some hypotheses have been advanced [10] Better understanding of the flow in rheometers is necessary to test these hypotheses in a meaningful way, and this model provides a means to develop that understanding. At this point, two types of rheometer were simulated: parallel plate, such as the BTRHEOM, and coaxial (Figure 2), such as the BML or CEMAGREF. The descriptions of these rheometers are given in ref. 10. Some preliminary results showed that if the inner cylinder of the coaxial rheometer does not have fins or serrations, the coarse aggregates have a tendency to move outward as a result of the presence of shear gradients. This implies that the torque measured is really the torque generated by a coupling of the rotor to the mortar phase and not to the concrete. This result is in agreement with the general knowledge that the rotating cylinder needs to be strongly coupled with the concrete and therefore it should not be smooth but have fins.
Figure 1: Simulation of SCC flow between two rebars: A) aggregate diameter is ½ and B) aggregate diameter is 1/5 of the distance between the rebars. The rebars (white circles) are pointing into the figure and the flow is in the direction of the arrow.
Figure2 : Simulation of a coaxial rheometer. The inner cylinder is rotating while the outer cylinder is fixed.
Obviously, a more realistic simulation of a concrete needs to take into account that the aggregates are not spheres or ellipsoids. Collaboration with the Federal Highway Association [11] allowed us to obtain X-Ray tomography-based images of aggregates (Figure 3). These images were incorporated into the code and can now replace the spheres in our simulation.
Figure 3: Reconstructed X-ray computed tomograph of a concrete specimen (270x270x270 pixels) (about 108 mm x 108 mm x 108 mm). This cube has been "cut" from a larger image, so that the flat faces of the aggregates at the surface are artificial.
Next: Conclusion Up: Main Previous: Main