12
1 The Experimental Approach in Aerodynamic Design
for aerodynamic installations (see Sect. 3.2). In a lesson learnt from the sonic boom
of the Concorde, for a supersonic aircraft, the origin and control of shock waves that
propagate to the ground is a critical issue.
1.4 Some Constraints for Wind Tunnel Test
There is a wide range of experimental means to carry out measurements required to
validate the numerical techniques mentioned above, but their degree of reliability,
ease of use or accuracy are extremely variable. In practice, it is often very challenging to perform several types of measurements on the same model or in the same
facility or laboratory. Indeed, the use of sophisticated techniques required to perform
these measurements is so complex that it may necessitate highly specialised teams
of experimentalists working on specific installations. Thus, for the purpose of validating the numerical models, it is often necessary to rely on several experimental
campaigns, where first the aerodynamic forces, pressure and temperature distributions are measured, followed by more complicated measurement of the average and
turbulent flow-fields from another technique and possibly the density from a third
source in the case of compressible flows.
Nevertheless, advances in measurement techniques over the last 40 years, including the advent of laser-based optical methods, have made a real breakthrough in our
ability to analyse complex flows containing shock waves, strong expansions, shear
layers, vortex organisations and recirculation regions.
Running experiments in a wind tunnel (or any similar installation) is also a simulation, here similar environment in which the vehicle will operate is recreated, the
model being tested being often much smaller than the full scale vehicle. As mentioned above it is of great importance to reproduce the viscous effects quantified by
the Reynolds number, which should be identical for the small scale model and the
real vehicle. Other similarity parameters such as the Mach number for compressibility effects, the Prandtl number for heat conduction, the Lewis number for mixing
species, the Knudsen number for rarefaction effects, must be conserved as well (see
Sect. 2.4). In hypersonic studies, it is recommended to use a gas with the same
composition and thermodynamic properties as the actual gas. Not all characteristic
numbers can be reproduced simultaneously and various facilities have to be used to
simulate the different flow conditions encountered by the vehicle, some conditions
being impossible to reproduce in ground facilities.
During most wind tunnel tests particular attention is paid to the Reynolds and
Mach numbers, the gas (in this case air) being the same in the practical application
and in the wind tunnel. However, during validation studies focusing on the physics of
complex aerodynamic phenomena, the Reynolds number is not as critical as it is proclaimed. Indeed, for a well-established turbulent regime, the characteristics of a flow
are almost independent of the Reynolds number. To a large extent, the main influencing parameters are boundary-layer properties just prior to separation (velocity
1 The Experimental Approach in Aerodynamic Design
for aerodynamic installations (see Sect. 3.2). In a lesson learnt from the sonic boom
of the Concorde, for a supersonic aircraft, the origin and control of shock waves that
propagate to the ground is a critical issue.
1.4 Some Constraints for Wind Tunnel Test
There is a wide range of experimental means to carry out measurements required to
validate the numerical techniques mentioned above, but their degree of reliability,
ease of use or accuracy are extremely variable. In practice, it is often very challenging to perform several types of measurements on the same model or in the same
facility or laboratory. Indeed, the use of sophisticated techniques required to perform
these measurements is so complex that it may necessitate highly specialised teams
of experimentalists working on specific installations. Thus, for the purpose of validating the numerical models, it is often necessary to rely on several experimental
campaigns, where first the aerodynamic forces, pressure and temperature distributions are measured, followed by more complicated measurement of the average and
turbulent flow-fields from another technique and possibly the density from a third
source in the case of compressible flows.
Nevertheless, advances in measurement techniques over the last 40 years, including the advent of laser-based optical methods, have made a real breakthrough in our
ability to analyse complex flows containing shock waves, strong expansions, shear
layers, vortex organisations and recirculation regions.
Running experiments in a wind tunnel (or any similar installation) is also a simulation, here similar environment in which the vehicle will operate is recreated, the
model being tested being often much smaller than the full scale vehicle. As mentioned above it is of great importance to reproduce the viscous effects quantified by
the Reynolds number, which should be identical for the small scale model and the
real vehicle. Other similarity parameters such as the Mach number for compressibility effects, the Prandtl number for heat conduction, the Lewis number for mixing
species, the Knudsen number for rarefaction effects, must be conserved as well (see
Sect. 2.4). In hypersonic studies, it is recommended to use a gas with the same
composition and thermodynamic properties as the actual gas. Not all characteristic
numbers can be reproduced simultaneously and various facilities have to be used to
simulate the different flow conditions encountered by the vehicle, some conditions
being impossible to reproduce in ground facilities.
During most wind tunnel tests particular attention is paid to the Reynolds and
Mach numbers, the gas (in this case air) being the same in the practical application
and in the wind tunnel. However, during validation studies focusing on the physics of
complex aerodynamic phenomena, the Reynolds number is not as critical as it is proclaimed. Indeed, for a well-established turbulent regime, the characteristics of a flow
are almost independent of the Reynolds number. To a large extent, the main influencing parameters are boundary-layer properties just prior to separation (velocity