Loading...
Please wait, while we are loading the content...
Large-Scale Particle Image Velocimetry on a Full-Scale Pool Fire
| Content Provider | Semantic Scholar |
|---|---|
| Author | Horváth, István Beeck, Jeroen Van Buchlin, Jean-Marie |
| Copyright Year | 2012 |
| Abstract | This study is a presentation of Large-Scale Particle Image Velocimetry (LS-PIV) on a Full-Scale Pool Fire (1 MW). The investigated pool fire was used for car fire simulations in car parks[4]. Small-scale measurements were conducted in order to support the full-scale measurements. The full-scale and the smallscale setups were situated on the premises of WarringtonFireGent NV and at the Von Karman Institute, respectively, in Belgium. The main objective is to describe the applicability and restrictions of PIV technique to full-scale fire dynamics. The application of PIV systems for measurements where fire is in the field of interest and consequently the ambient light is inevitable is the challenge here. The problems related to the new demands are introduced, and possible solutions and advices are given. Firstly, a small-scale fire torch (60 mm x 100 mm) is used in order to recognize all the possible additional elements and particular settings with respect to standard PIV. By cutting on the integration times (5μs), applying Extreme Digital Range option ([8] / EDR = 1) and posterior background subtractions ([2] / Subtraction). The improvement is significant. The Average SN Ratio increased from 0.94 to 3.35 while the bad vector percentage dropped from 87.9% to11.1% in the presented particular case. Secondly, further information is gained by trials of the new findings on the full-scale fire (2000 mm x 3500 mm). Real-scale pool fire tests allowed to obtain the maximum applicable pixel displacement for large scale PIV on the turbulent structures of the flame. Its value, normally 8 pixels for standard PIV, decreased to 3 pixels using WIDIM for the presented case. This demands shorter separation times and thus results in a lower dynamic range for the velocity. The reason is the fact that turbulent structures are used instead of the tracer particles, whose shape can change in time as well. Finally, by combining the smalland full-scale experiences, successful LS-PIV measurements are carried out on the full-scale fire. The acquired midline flame velocity results, despite of the disturbance of the ventilation flow, are in agreement with the results of McCaffrey’s model [5] when both IstInfo and Signal-toNoise ratio filtering [1] are applied during the post-processing. This agreement proves the applicability of Large-Scale PIV (LS-PIV) to turbulent flame structures and validates the semi-empirical model of McCaffrey. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://ltces.dem.ist.utl.pt/lxlaser/lxlaser2012/upload/227_abstract_soisoj.pdf |
| Alternate Webpage(s) | http://ltces.dem.ist.utl.pt/lxlaser/lxlaser2012/upload/227_paper_soisoj.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
