Abstract
This State-of-the-Art report (STAR) gives an overview over factors influencing particle packing and porosity. An overview of methods for measuring and modelling particle packing and porosity is also made. Improved particle packing of the concrete matrix (i.e. particles < 125 μm) will reduce porosity and permeability of concrete and thereby improve durability of concrete. The findings are based on literature reviews.
It appears like the fineness of the particles is determining the size of the pore openings governing the pore connectivity and thereby permeability if they are added in large enough quantities to disperse the coarser ones, which is obvious from basic particle packing theory. An excess of fine particles dispersing coarse ones (e.g. high dosage of silica fume relative to cement) can dominate the permeability all together, without being the result of a refined complex particle packing. For more normally distributed multiple particle compositions, a wider particle size distribution is more beneficial to obtain optimize particle packing.
It is also demonstrated that it is not only the initial packing of particles, but also the increased volume of their solid reaction products that reduces permeability. Even particles considered by many to be inert (like limestone powder) will react when given sufficient reactants (e.g. calcium alumina hydrates from pozzolanic reaction of fly ash) and theoretically contribute positively to reduced permeability.
Thus, making smallest possible reacting particles would give the ultimate low pore opening size. The challenge is to make them in the most economical way (grinding is expensive and limited to a certain size) and at the same time not let them hamper workability by being too reactive in the fresh state. Thus, the fine tail of particle size distributions of e.g. ternary cement should consist of slowly reacting particles; for instance nano-sized, precipitated calcium carbonate, clay calcined below sintering temperature easily ground to 3 μm or perhaps fine recycled glass powder. Further research is recommended along these lines.
It is also speculated whether initial Ettringite formation in fresh cement paste may hamper particle packing initially or not, and search for other setting regulators than gypsum may be a way to find this out. A gypsum free system may also allow high temperature curing and enable increased productivity. Future research should clarify this matter.
The most feasible way of determining particle packing in the fresh state seems to be a method called “centrifugal consolidation”, due to its simplicity, rapid result and operator independence.
Computer programs for calculating pore connectivity and permeability of hydrating cement paste is also reviewed and it seems that all programs suffer from the same problem; their resolution versus computing time. Many works with pixel values in excess of 1 μm and in concrete matrix there are many pores below this value that will not be predicted.
The most important parameters governing permeability and percolation of a concrete matrix seems to be pore connectivity and critical pore opening. There are indications showing that a concrete matrix will depercolate at a total porosity of 18-20 vol% independently of being obtained by low water-to-cement ratio, particle packing, degree of hydration etc.
When permeability of concrete matrix is lower than the whole concrete, it is pointed out that the permeability of the actual normal density aggregate may be higher than a well cured cementitious paste, and that the permeability increase observed for concrete not necessarily is due to a more porous interfacial zone between aggregate and matrix that many authors seem to focus on.