This work deals with the topic of dynamic calibration of pressure sensitive paint (PSP) using methods of system identification in z-space, suitable for discrete time analysis. The aim is to be able to reconstruct time varying pressures from (phase and amplitude distorted) measurement data. The calibration method is useful when a well known pressure source is available, such as moving shock in a shock tube. 

In this experiment ruthenium based polymer/ceramic pressure sensitive paint (PC-PSP) has been used to study the pressure inside a y-junction at pulsating flow conditions. Pressure has been measured using the intensity based method and through phase locked averages. The aim has been to investigate the potential of PC-PSP at mass flows and pulse frequencies typical of those in the exhaust manifold of internal combustion engines. 

When using pressure sensitive paint under unsteady conditions in low-speed applications, the signal-to-noise ratio is usually low and may hinder the proper evaluation of the acquired data. Here, we propose a new noise-filtering scheme that is based on singular value decomposition. As a test case, we evaluate the fluctuating pressure field due to unsteady vortex shedding on the side of a square cylinder. The proposed scheme resulted in a reduction of pixel noise of the order of two magnitudes which made it possible to obtain results regarding the spatial form of flow structures as well as the shedding frequency.

The pressure and velocity fields in a y-junction of a square (40×40 mm²) cross-section channel were investigated during pulsating flow. One of the sides of the channel was covered with fast responding pressure sensitive paint (PSP) whereas the velocity field at the channel center parallel to the PSP surface was measured using particle image velocimetry (PIV). The flow conditions, in terms of mass flow rate and pulsation frequency, were selected to resemble the flow inside an exhaust manifold of a small internal combustion engine, although the gas was at room temperature. The mass flow was varied between 10 and 130 g/s with pulsations between 0 and 80 Hz. For both the PSP and the PIV measurements images were acquired unsynchronized to the pulses using a high-speed camera and phase averages were formed a posteriori. The use of PSP together with PIV demonstrates how the two techniques can be used to verify and complement each other, PIV excelling at the lower mass flow rates and PSP at the higher. It is shown that the signal-to-noise ratio for PSP at low velocities can be enhanced using a technique based on singular value decomposition.