Experimental study of particle losses close to the entry of thin-walled sampling probes at varying angles to the wind

Citation
A. Sreenath et al., Experimental study of particle losses close to the entry of thin-walled sampling probes at varying angles to the wind, AEROS SCI T, 35(3), 2001, pp. 767-778
Citations number
20
Categorie Soggetti
Mechanical Engineering
Journal title
AEROSOL SCIENCE AND TECHNOLOGY
ISSN journal
02786826 → ACNP
Volume
35
Issue
3
Year of publication
2001
Pages
767 - 778
Database
ISI
SICI code
0278-6826(200109)35:3<767:ESOPLC>2.0.ZU;2-P
Abstract
This article summarizes the results of an extensive experimental study of s ampling losses in thin-walled probes at various values of velocity ratio R and the probe orientation with respect to the freestream. The purpose of th is study was to gain insights into the complex interaction of various param eters that influence sampling losses and the consequent effect on the overa ll sampling efficiency. A 0.635 cm diameter sharp-edged tube was mounted in a small wind tunnel where the freestream velocity could be varied over a w ide range of values. Polydispersed spherical glass beads were used as the t est aerosol. The number concentration and the particle size distribution we re measured using the aerodynamic particle sizer (APS 3310). The sampling e fficiency was determined as a function of orientation for a range of partic le sizes (or Stokes number). By using an existing model to predict the aspi ration efficiency for thin-walled probes, the sampling losses could be isol ated from the sampling efficiency. In this manner a new empirical model was developed to predict the losses as a complex function of Stokes number, sa mpler orientation, and velocity ratio. The losses appear to be influenced b y particle inertia, impaction, gravitational settling in the boundary layer developing inside the thin-walled probe, and vena contracta or flow recirc ulation loss near the entry. It was evident from the results that these los ses are strongly influenced by the Stokes number and sampler orientation. T he losses also increased strongly with increasing value of velocity ratio f or all orientations.