Entrainment of oil droplets at the back of mist filters

• Applicant: Professor Dr. Gerhard Kasper
• Subject Area: Mechanical Process Engineering
• Term: from 2012 to 2017
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 227377638

Entrainment of previously deposited oil aerosol from mist filters occurs as a result of gas flow through the partially saturated media. Entrainment generates a broad droplet spectrum (0.1 - 2000 µm). Preliminary work by the PI suggested two possible entrainment mechanisms, one for oleophilic media (bursting bubbles in the oil film on back wall) and for oleophobic media (droplet blow-off). On basis, firstly, measurement techniques were developed to determine entrainment rates and entrained drop sizes in two relevant size ranges (0.5 - 10 µm and 170 - 2300 µm). Additionally, imaging techniques were developed to observe drop fragmentation and bubble formation. Finally a cumbersome, but useful off-line technique for measuring in the gap (10 - 200 µm) was implemented. A combination of these measurement methods enabled the recording of complete spectra. These results confirmed that the above mentioned size ranges should suffice for interpretation of entrainment kinetics. Another part of the work focused on obtaining an understanding of the relevant entrainment mechanism(s) as function of media wettability and operating conditions (e.g. aerosol mass rate and filtration velocity) under steady state conditions. By combining entrainment rate measurements and imaging techniques it was possible to correlate entrainment and drainage processes. Surprisingly, direct blow-off from oleophobic filters does not contribute noticeably to overall entrainment, except for an initial burst which is related indirectly to blow-off. However, a second mechanism was discovered for oleophobic media which is based on fragmentation of bubbles on the surface of drops. In case of wettable media, bubble bursting seems to be the dominant mechanism. Based on these findings, the following work program and strategy was developed for the next project period: Foremost, the hypothesis needs to be confirmed that fragmentation of bubbles is the predominant entrainment mechanism for glass fiber media in steady state. This will be done by correlating entrainment rates with bubble formation rates under equal operating conditions. It requires more extensive entrainment rate measurements, as well more quantitative data on bubble bursting rates and resulting fragment size spectra. Second objective is to gain insight regarding where bubbles form, as well as any preferential points of bubble formation on the filter. This will be done by using additional types of media, as well as structured surfaces (e.g. foam, grid). These tests will be augmented by experiments of bubble formation on free liquid surfaces in a separate set-up with more control over bubble sizes and rates. These data are needed for the interpretation of data from real filter media. Overall, this should lead to a well-rounded understanding of entrainment mechanism(s) from glass fiber media, as well as the dependence of entrainment rates on operating conditions, and contribute to strategies for the reduction of entrainment.

• DFG Programme: Research Grants
• International Connection: Australia
• Cooperation Partner: Professor Benjamin James Mullins, Ph.D.