Core sands for blowing processes, belong to these sands in which small amount of the applied binding material has the ability of covering the sand matrix surface in a way which - at relatively small coating thickness - allows to achieve the high strength. Although the deciding factor constitute, in this aspect, strength properties of a binder, its viscosity and ability to moisten the matrix surface, the essential meaning for the strength properties of the prepared moulding sand and the mould has the packing method of differing in sizes sand grains with the coating of the binding material deposited on their surfaces. The knowledge of the influence of the compaction degree of grains forming the core on the total contact surface area can be the essential information concerning the core strength. Forecasting the strength properties of core sands, at known properties of the applied chemically hardened binder and the quartz matrix, requires certain modifications of the existing theoretical models. They should be made more realistic with regard to assumptions concerning grain sizes composition of quartz sands and the packing structure deciding on the active surface area of the contacts between grains of various sizes and - in consequence - on the final strength of cores.
The results of investigations of the influence of the matrix grain sizes on properties of cores made by the blowing method are presented in the hereby paper. Five kinds of matrices, differing in grain size compositions, determined by the laser diffraction method in the Analysette 22NanoTec device, were applied in investigations. Individual kinds of matrices were used for making core sands in the Cordis technology. From these sands the shaped elements, for determining the apparent density of compacted sands and their bending strength, were made by the blowing method. The shaped elements (cores) were made at shooting pressures being 3, 4 and 5 atn. The bending strength of samples were determined directly after their preparation and after the storing time of 1 hour.