QLS measures elastically scattered light by small particles. In general, these particles have to be added to the flow. They have to fulfill a number of boundary conditions in order to be well suited for QLS experiments.
The light is scattered by ensembles of individual scatterers. The recorded signal in each cell of the receiving optics will always be a superposition of a large number of individual scattering processes.
QLS measurements can only be executed, when two flow components can be separated. Only in these cases it is possible to seed one flow component and to leave the other component unseeded.
The signal in the measurement volume is then proportional to the mass concentration of the mixing partners.
Both flow components are seperated before they enter the mixing section. One component is seeded with small particles. The receiving optics measures the dilution process downstream of the injection hole.
The acquired raw images in QLS measurements include a large number of influencing parameters. In some cases it is possible to reduce the influence or even compensate it. In others more or less sophisticated correction algorithms have to be applied in order to obtain accurate concentration data.
But even with very sophisticated post processing it is not possible to express the concentration value absolutely. This is why the intensity value at the jet entering location is set to 100% and the intensity values upstream (recirculation) and downstream (dilution) are correlated to this value.
Using this correlation one receives the concentration field as values between 0 and 1, where the value 1 is reached at the jet centerline at the point, where the jet enters the mixing field.
The most prominent influencing parameter is the scattering direction. In laboratory experiments have been carried out showing that an acquisition under 10° forward scattering angle (refering to normal view) causes an increase in intensity of 50%. It is therefore necessary to have reliable datasets for this calibration
The incoming laser light is generally stronly polarized. It is well known that Mie scattering is largely polarization conserving. Although the scattered light is elliptically polarized, the amplitude of this effect was found out to be lower than 10%.
In this case the influence can be redurced experimentally by insertion of a lambda fourth plate into the generating optics, so that the laser light becomes circularly polarized. In this case the polarization influence vanishes.
By traveling through the measurement volume the light sheet intensity is decreased by each scattering process. In practical applications and under appropriate seeding conditions, extinction values of 10-15% are quite common. There is no simple way to compensate this effect, because it is field dependent and therefore the local particle concentrations have to be taken into account.
An accurate correction scheme has been setup by Voigt (see Voigt and Schodl 1997). The integral intensity loss and the raw images are used for a numerical correction based on "shooting".
The algorith converges rapidly and works up to intensity losses of 80%.