The usual test method is the Charpy impact test. It measures the energy to fracture a pre-notched standard sized specimen.  The test is much quicker than a tensile test (so less time for the specimen to warm or cool towards room temperature), and there is less scatter in the data due to the pre-Vee-notched specimen.
The fracture energy (either in Joules or ft-lbs) is then plotted vs. temperature to determine the transition temperature.
Sometimes called the Charpy V-notch (CVN) test.
Lots of data, the test method is 50-60 years old.
It is also used to study the effect of microstructure, alloy composition, etc. at room T.

See ASTM E23-02a Standard Test Methods for Notched Bar Impact Testing of Metallic Materials
http://www.astm.org/DATABASE.CART/REDLINE_PAGES/E23.htm

Read my response to Thread725-59759. It should answer any additional questions.


                                             maui

When you report the B/B temperature, make certain that you also report the criterion you used.  There a number of different ones.

Many testing houses can perform the test. It's advantageous to use a fast impact test (high strain rate) which includes Charpy. Ductile/brittle transition is for a very specific strain rate and specimen size and shape.

In its theory the transition occurs when the strain, inhibited by lower temperature, can not occur rapidly enough to accommodate the applied stress. Fracture may occur after a finite amount of initial strain, however, depending upon what you decide to accept. Look up general test specifications at:

          http://www.astm.org

Fracture occurs when the part fails through rapid crack propagation fully across any given cross-section. The notch pre-determines the "crack" size and geometry. The fracture surface will be jagged and consist of a great deal of exposed, granular facets.