Investigation on reliability of particle packing models for aggregates

Abstract: Concrete is the most used manmade material in the world. Concrete is a composite material mainly made of four basic ingredients: cement, sand - or fine aggregate, coarse aggregate, and water. Fine and coarse aggregates occupy 60 % to 80 % of the concrete volume. The packing density - that is, the ratio of solids to void - of its components affects the concrete’s behaviour: a higher amount of packing of aggregates leads to minimizing the amount of cement needed in volume of concrete. Since it is usually desirable to achieve as high packing density as possible in order to minimize the inter-particle voids between the constituents of concrete, the density must be accurately calculated. A high packing density thus leads to a smaller need for cement, which is desirable both for the sake of mechanical property and of the environment. Mathematically, it is possible to calculate the packing density of mono-sized spherical materials.
A number of particle packing models have been developed over the past 80 years. However, no existing particle packing model today can accurately predict the packing density of different aggregate mixes. Therefore the purpose of this thesis is to examine the reliability and accuracy of particle packing models in comparison with actual data obtained in the laboratory. Since the particle packing models can be used as a part of concrete mixture design, the thesis can provide a foundation for developing a stronger and more durable concrete mixture, while taking environmental aspects into account.
Hence, in this study, packing densities of seven mixes of aggregate were attained in a laboratory, using the loose packing method, and were then compared to values suggested by two particle packing models commonly used in Scandinavia: Modified Toufar Model and 4C.
The comparison between the models' and the laboratory results has shown that 4C proved more accurate in determining packing density for certain mixtures, especially within the span of 40 to 60 % fine aggregates of the total volume. However, Modified Toufar was most successful overall in predicting the packing densities, resulting in a smaller total mean difference from the laboratory test results. It was also found that, when using 4C for predicting the packing density of a concrete mixture using the loose packing method, Mu-value 0.07 gave the best fit for the materials used in this study. This implicates that the recommendation in the 4C manual for choosing Mu-value 0.07 for Danish material, can be applicable for Swedish material as well.
Also, it was found that 4C, when cubic crushed material was used as fines in a mixture with naturals as coarse aggregates, rendered unrealistic results with a local minimum packing density. In order to come to terms with the reason for these deviations, and to provide more solid results with the 4C model, further research, as well as development of the software, might be of interest in the future.