Determination of statistical strength in tensile testing of cured molding sands

One of the main physical and mechanical properties that determine the quality of molding and core sands is the ultimate tensile strength in the cured state. Many years of experience in its application in production and scientific research should correspond to the reliability of the obtained indicators and their low scatter when obtained according to GOST 23409.7-78 in the Russian Federation and DIN, ASF abroad. However, the practice of applying existing standards allows to state the presence of serious shortcomings of this method, manifested primarily in a large scatter of the measured values of the strength of samples prepared and tested under the same conditions. It is shown that when determining the tensile strength of hardened molding and core mixtures, 30-40 % of the tests must be repeated, since the requirement for a 10 % scatter of indicators from the average value of three tests is not met. If this requirement is met, then the average values of the obtained indicators differ by 50-60 %. The method of determining the strength value by the average of three indicators does not have a physical justification for the destruction process. During testing of these samples, cracking sounds can be heard, which indicates that when a load is applied to the sample, the process of destruction began almost immediately, but the sample itself is not destroyed immediately, but after some of the bonds between the grains of low-strength refractory filler have been destroyed. Methods are proposed for determining the statistical and true strength of molding and core mixtures, taking into account the physical processes occurring in the test samples when a destructive load is applied to them. The accuracy of determining the strength indicators of mixtures in the cured state by statistical strength is significantly higher than by the existing method. It is shown that the standard deviation determined by the existing method is 2–3 times higher than by statistical and true strength, which indicates a smaller scatter of strength indicators.