The analysis of turbulent boundary layers (TBLs) has been a cornerstone of the modern conception of fluid mechanics, particularly with respect our understanding of the nature of turbulent shear flows and their mixing behaviour. In recent decades, the importance of the turbulent/non-turbulent interface (TNTI) to characterisation of the entrainment and detrainment of flow into and out of the turbulent boundary layer has increasingly come to the forefront of experimental research. The highly sensitive nature of turbulent boundary layers demands careful conditioning of flow conditions to avoid the introduction of bias into the data; the ability of upstream conditions to persist far downstream is well-documented. To inform the design of experiments in the Atmospheric Wind Tunnel Munich (AWM) and investigate the existence and origins of irregularities observed in the boundary layer on the sidewall, a series of measurements were conducted near the start of the test section. These motivated the development of a large, low-cost multilevel calibration targets to enable simultaneous aligned measurement of fields-of-view spanning up to 1.1m×0.4 m. These large-scale measurements and additional smaller-scale tests were used to identify the footprints of vortices propagating from the flow turning vanes along the sidewall and inform a renovation of the flow contraction to remove other sources of inhomogeneity. A significant challenge in observing the TNTI and macroscopic entrainment and detrainment behaviour is accurate detection and segmentation. To conduct flow tagging of the TBL without compromising access to full-field velocimetry, a novel solution of Pyrromethene 567 (P567) in Di-Ethyl-Hexyl-Sebacate (DEHS) developed at the Von K´arm´an Institute was repurposed. Although originally developed for fluorescent PIV measurements, the seeding particles were introduced into the boundary layer and used as a passive tracer with which to perform image segmentation and separate flow originating within the TBL from the freestream. This technique was used to obtain intermittency profiles, p.d.fs of macroscopic entrainment and detrainment behaviour, and to perform zonal decomposition of the flow. Analysis based on this new dataset indicated the presence of bias in 2D characterisation of the topology of entrained and detrained areas due to the fundamentally three-dimensional nature of TBLs. Finally, an experiment to extend the P567-based flow segmentation in conjunction with a time-resolved SPIV measurement in the cross-stream plane was designed. It was aimed at obtaining a 3D reconstruction of the TBL through the application of Taylor’s hypothesis with access to full-field 3-component velocimetry. A more limited implementation based on a local-seeding only approach was successfully used to reconstruct the TBL over a domain spanning approximately 200δ×1.5δ×2.1δ, albeit without velocimetry outside the boundary layer. However, this dataset was applied to analyse the extent to which measurements of the TNTI and macroscopic entrainment and detrainment obtained from 2D measurements are biased. Preliminary results indicate that there is significant bias introduced by measurements in the classical streamwise wall-normal plane, with overprediction of both entrainment and detrainment.     «

The analysis of turbulent boundary layers (TBLs) has been a cornerstone of the modern conception of fluid mechanics, particularly with respect our understanding of the nature of turbulent shear flows and their mixing behaviour. In recent decades, the importance of the turbulent/non-turbulent interface (TNTI) to characterisation of the entrainment and detrainment of flow into and out of the turbulent boundary layer has increasingly come to the forefront of experimental research. The highly sensit...     »