TPT January 2012

A rticle

For the structures with abnormally coarse grains circled with sufficiently homogeneous fine grains, there is a special standard, ASTM Е 930 Methods for Assessment of the Largest Observed Grains (ALA), which has a number of drawbacks and practically is not used in industrial practice. In the present-day practice of making products from high-alloy steels and alloys, one of the main requirements determining quality of the finished products is the grain size determined by the methods established in the existing standards. The following standards were most commonly used at the end of the last century: • ASTM E l12-95: Standard Test Methods for Determining Average Grain Size • DIN 50601: Determining Ferrite and Austenite Grain Size in Iron Base Steels and Materials • ISO 643: Steel. Micrographic Determination of the Apparent Grain Size • GОSТ 5639: Steels and Alloys. Methods for Determining Grain Size It should be noted that all ASTM standards set narrow scopes of their competence. It ensures determining grain size in corresponding (though rather conventional) units with a rather high reliability. As regards standard GOST 5639, its authors’ attempt to establish a universal normative document has resulted in a number of uncertainties and, in some cases, in impossibility of a reliable grain size determination in steels prone to anisotropy, namely the steels for which anisotropy is the main quality index. This is why there were cases of rejection of this standard application in industry. The most serious drawbacks of GOST 5639 include the following: • setting limits of grain quantity in 1mm 2 for each grain number • introduction of a method for assessment of anisomerous structures • its obligatory use raised to the rank of a law All three standards contain different quantities of whole numbers conventionally defining grain sizes. For all that, ASTM Е112 permits assessment in fractional numbers (eg G8.5). It should be pointed out that average grain sizes with the same numbers are identical in the three standards. Standards DIN 50601 and GOST 5639 use a metric system in which number of grains in 1mm 2 is determined and the average diameters approximately correspond to numbers G0 tо G14 in ASTM E112 in which another size scale in inches is additionally given. Use of the metric system in European standards has brought about a shift of the standard sequence into the field of negative numbers having no physical sense. Common disadvantages of all three standards include conventionality of units (ie numbers) which complicates their conversion into the commonly used units (millimetres) in design calculations. For such method standard as GOST 5639, its obligatory use is incorrect because it includes provisions allowing double interpretation of the same quantity.

GOST 5639 contains a number of intolerable errors which were discussed in detail in [4-6] . Determination of average grain diameter by the method of grain boundary intersection (it is unclear why it was called “conditional”) is analogous to that of E112 but numerical value d L (GOST 5639) and l - (ASTM Е112) somewhat differ due to the difference between British and metric systems of units. Major doubts are raised by the method of counting the number of grains in 1mm 3 in GOST 5639, ie counting an average number of grains in the specimen volume not specified in any one of the standards under consideration. The American standard specifies determination of an average number of nonequiaxed grains in 1mm 2 and GOST 5639 in 1mm 3 which is totally baseless and can only be done with the use of stereological reconstruction. It would be quite another matter if it were a question of area. Then, according to [5] , an equigranular in volume structure (ie a structure consisting of grains of a same size) for which D max = D av = D min gives distribution containing up to 28% of grains sizing from D min → 0 to D max in a random secant plane. For comparison, Tables 1 and 2 list grain structure parameters and methods for their determination in accordance with the three most widespread standards to demonstrate basic differences between them. The final objective of development of the technology for manufacture of products of special steels and alloys is formation of a specified uniform grain size which in its turn ensures required design material strength. As such structures are mostly assessed in practice by the method of visual comparison with photographic references from the functional standards and in conventional units (numbers), such assessment is subjective, open to serious measurement errors and methods of assessment of grain size unhomogeneity are unavailable at all. As the analysis of the existing standards shows, except the chord method, they do not take into account real distribution of the grain characteristics (chords, plain section diameters, diameters in a volume). Therefore, it is inexpedient to use it for assessment of high-duty products in which grain size is the main quality index, and as SA Saltykov affirms in his book, “Standardised methods of metallographic analysis… are mostly semiquantitative methods of number-based assessment. Its main disadvantage is its subjectivity and the resulted low accuracy and reliability”. An objective assessment of grain structures in metals and alloys can be done with the help of the function of distribution of three- dimensional objects and its quantitative characteristics which can be realised through a stereological reconstruction of the actual structure images visible in microsections and by the use of computerised methods of primary data processing [5] . The known quantitative methods in the standards of developed countries (ASTM Е112, DIN 50601, etc) allow to measure grain size with identical errors.

117

www.read-tpt.com

J anuary 2012

›