AGARD-B is a standard wind tunnel model (calibration model) that is used to verify, by comparison of test results with previously published data, the measurement chain in a wind tunnel. Together with its derivative AGARD-C it belongs to a family of AGARD standard wind tunnel models. Its origin dates to the year 1952, and the Second Meeting of the AGARD Wind Tunnel and Model Testing Panel in Rome, Italy, when it was decided to define two standard wind tunnel model configurations (AGARD-A and AGARD-B) to be used for exchange of test data and comparison of test results of same models tested in different wind tunnels. The idea was to establish standards of comparison between wind tunnels and improve the validity of wind tunnel tests. Among the standard wind tunnel models, AGARD model configuration B (AGARD-B) has become by far the most popular. Initially intended for the supersonic wind tunnels, the AGARD-B configuration has since been tested in many wind tunnels at a wide range of Mach numbers, from low subsonic (Mach 0.1), through transonic (Mach 0.7 to 1.4) to hypersonic (up to Mach 8 and above). Therefore, a considerable database of test results is available.
AGARD-B is a body-wing configuration. All its dimensions are given in terms of the body diameter "D" so that the model can be produced in any scale, as appropriate for a particular wind tunnel. The body is an 8.5 diameters long solid of revolution consisting of a 5.5 diameters long cylindrical segment and a nose with a length of 3 diameters and having a local radius defined by the equation .
The wing is a delta in the form of an equilateral triangle with a span of four body diameters. Wing section is a symmetric cylindrical arc with a relative thickness t/c of 4%. Leading and trailing edges of the wing should be rounded with a radius equal to . However, this specification is unclear. It is obvious that the specified radius can not be applied near the wingtips, or large deformations in the plan form of the wing would occur. In the past, this part of the specification was interpreted in different ways by model designers which led to small differences in shapes of the tested models. The recommended solution is to have the leading- and trailing-edge radii of at the theoretical root chord and to decrease the radii towards the wing tips proportionally to the local chord.
A support sting to be used with the AGARD-B model was defined as well. The initial specification of the model called for a sting having a diameter of and a length of . In the revised specification the length of the sting was changed to in order to reduce sting interference, but at that moment a number of wind tunnel tests had already been made. Therefore, published test results for the AGARD-B models do not all correspond to theoretical model configuration.
The drag characteristics of the AGARD-B model were found to be somewhat sensitive to the boundary layer transition on the model. In order to reduce the scatter of results, in some wind tunnel facilities the model was tested with boundary layer transition trips near the leading edges of the wing and the nose of the body. On the other hand, a number of wind tunnel tests was made without fixed transition. Drag results with and without the fixed boundary layer transition differ, which should not be neglected when comparing test results from different wind tunnel laboratories.
In some wind tunnel laboratories, AGARD-B was tested in non-standard configurations, e.g. as a half-model (half-span model).
Some free-flight tests of the AGARD-B model were performed. For these tests, the standard geometry was modified by adding, at the rear end of the body, two triangular vertical stabilisers, one on the ventral and one on the dorsal side of the body. Size of the vertical stabilizers was 50% of the wing size, i.e. their span was .