(Received 4 May 2005; accepted 24 October 2005)
| ||Format||Pages||Price|| |
|10||$25||  ADD TO CART|
During the past decade, greatly increased in-service demands on finished products such as gears, bearings, and springs have inspired a wish by steel makers to guarantee the highest level of cleanness in their steels. In order to achieve this, it has been necessary to develop methods of characterization of cleanness which can describe all the nonmetallic inclusion populations endogenous and exogenous contained in the steel. These methods are essentially based on quantitative metallography and ultrasonic tests. Statistical tools have also been developed to maximize the accuracy of measurements in relation to the time taken to make them, and to answer the key question: What is the minimum analyzed volume for which a measurement can be expected to yield a reliable estimate of a specific in-service property? For quantitative metallographic methods, smaller inclusion densities necessitate an increase in the surface area analyzed, and the study of a larger surface is time consuming. Extreme value analysis is a further method for prediction of the expected size of the largest inclusion in a volume. Often, however, the amount of material examined is not sufficient to assess the quality of the heat with complete certainty. Other tools that have been developed utilize high frequency ultrasonic tests over a frequency range from 10 to 100 MHz, which make it possible to detect inclusions with diameters ranging from 15 µm to 1 mm. In all cases, in order to obtain accurate estimates of the densities of nonmetallic inclusions, it is important to know the appropriate settings for the measurement method. In particular, it is very important to estimate the diagnosis error when we rate a product on the basis of measurements. For each of these inspection techniques, statistical models have been developed to assess the main statistical properties of the methods. The results are reported here. They give a basis for comparison of steel heats that takes account of knowledge of the confidence level of the various measurement methods.
Research Group Manager, CREAS, ASCOMETAL Research Center, HAGONDANGE Cedex,
Stock #: JAI14032