Open Access
Yan, Sihong
Graduate Program:
Aerospace Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
June 15, 2016
Committee Members:
  • Jose Palacios, Thesis Advisor
  • Engine Icing
  • Mixed-phase Ice Crystals
  • Impact
  • Luminescense
  • Melting Process
High altitude ice crystals have been recently discovered to be the cause of engine and heated probe icing over high humidity tropical regions. Ice accretion related to partially melted ice crystals was first discovered in 2006 and it is a threat to aviation safety. It is known that ice crystals without any water content do not accrete to surfaces. The classical frame icing theory involving super-cooled water droplets cannot explain the cause of icing inside turbofan engines flying at altitudes where there is no water content, since only fully glaciated ice crystal clouds exist. To understand the icing conditions and physical mechanism of engine icing, research projects like the High Altitude Ice Crystal (HAIC) international project are been conducted, and test facilities, like the National Research Council icing wind tunnel or the NASA Propulsion System Laboratory tunnel, have been constructed. The correlation between engine icing events and mixed-phase icing clouds that partially melt when ingested in an engine has been confirmed in these facilities. Despite the availability of facilities to reproduce the ice accretion events inside engines, fundamental testing of individual partially melted water droplets is not available and the validation of tools to predict partial melting of crystals is not possible. The study of physical processes involved in the partial melting of a single ice crystal can be divided into two parts. The first part is the impact dynamics of the single droplet, and the second part is the melting process of the frozen droplet. Attempts to characterize these two phenomena were conducted at the Adverse Environmental Rotor Test Stand Facility at Penn State. To quantify impact dynamics of ice crystals, high-speed video of single frozen water droplets impacting a surface was acquired. The frozen particles had a diameter ranging from 0.4 mm to 0.9 mm and impacted at velocities varying from 90 m/sec to 309 m/sec. The technique used to freeze the droplets and launch the particles against a surface is described. High-speed video was used to quantify the ice accretion area to the surface for varying impact angles (30⁰, 45⁰, 60⁰), and impacting velocities. An oxygen /acetylene cross-flow flame was used to partially melt the traveling frozen particles and it is also discussed. A linear relationship between impact angle and ice accretion is identified for fully frozen particles. The slope of the relationship is affected by impact speed. Higher impact angles closer to perpendicularity between the surface and the particle trajectory, e.g. 60⁰, exhibited small differences in ice accretion with varying velocities. Increasing velocity from 161 m/sec to 259 m/sec nearly doubled the ice accretion area at a shallower impact angle of 30⁰. The increase accretion area highlights the importance of impact angle and velocity on the accretion process of partially melted ice crystals. It was experimentally observed that partial melting was not a pre-requisite for accretion at the tested velocities when impact angles of 45⁰ and 30⁰ were used. The ice accretion due to impact was observed under five surface temperatures, -20°C, -15°C, -10°C, 0°C and 10°C. The influence of the surface temperature was qualitatively observed at an impact angle of 30°. The temperature varied from -15°C to 10°C, and a maximum area of ice accretion was observed at surface temperatures surrounding the freezing point of water. A second emphasis of the work was to correlate residence time requirements for the melting of frozen drops. To characterize the melting process of fully glaciated droplets, a luminescent technique was developed to measure the percentage of melting experimentally. Luminescent dye and high-speed camera visualization were used to monitor the partial melting state of an ultrasonically levitated frozen drop exposed to warm environments. Rhodamine B was dissolved (0.01% mass fraction) in the water used to create a droplet. The Droplet was placed at the node of the wave created by the acoustic levitator and frozen via convective cooling. When the cold air flow was turned off, the partial melting of the droplet began. Water droplets with a diameter ranging approximately between 300µm to 1800µm were tested. Four environmental melting temperatures were tested: 5°C, 15°C, 25°C and 35°C. The variation of percentage of partial melting of the drop with time was recorded. The correlation between the rate of melting, environmental temperature, and diameter of the frozen droplets is reported and discussed. It is confirmed that the time rate of melting is inversely proportional to the diameter of the ice crystals and directly proportional to the environmental temperature. An empirical fit to predict the percentage of partial melting with respect to temperature and droplet diameter was experimentally acquired. The models developed in this research can improve the understanding of the physics related to engine icing. In addition, several technologies developed during the effort can be applied to icing wind tunnel testing for the quantification of partial melting.