The relationship between material Brinell hardness and
material stress–strain behaviors has been extensively
studied. Both ISO/TR 10108 and GB/T 1172-1999 (Chinese
national standard) have elaborated a conversion of hardness
values and tensile strength values for steel [1,2]. In
2007, Janosec used the same strip steel as the measured
material both in a tensile test and a Brinell hardness test.
He gained the results of Brinell hardness, yield stress, tensile
strength and their ratio at the room temperature,
respectively [3]. In 2008, Tien discovered that the relationship
between the tensile strength and Brinell hardness value
of a material is very intensive [4].
Uncertainty evaluation of material Brinell hardness
measurement is usually a difficult problem. A probability
and statistics methodology (type A method of ‘‘Guide to
the expression of uncertainty in measurement’’ [5]) is
often used to evaluate the measurement uncertainty, for
example, in 2005, Herrmanzinn synthesized the uncertainty
of hardness measurements according to the error
propagation law [6]. Different from the conventional method
above, some new uncertainty evaluation methods have
been developed: Roberts applied Neural Network theory to