Initial development of ice crystal ice accretion at conditions related to turbofan operation at high Altitude

Initial development of ice crystal ice accretion at
conditions related to turbofan operation at high
Altitude

Ice accretion on external surfaces of aircraft is a widely recognised problem, but more recently identified problem of ice crystal ice accretion within aero-engine compressors
during flight through deep convection systems also represents a significant hazard and forms the motivation for the present work. The experimental studies targeting solid phase ice accretion are very limited due to the high wind tunnel facilities operational cost and safety concern for in-flight icing testing, which requires flight through severe weather conditions.

In this study, a small wind tunnel was established to simulate some of the conditions relevant to aircraft engine icing from ice crystals and explore the application of a model for the initiation of ice accretion. In this facility, liquid nitrogen was used to freeze liquid water droplets generated using an ultrasonic nozzle. The liquid nitrogen section reduces the droplet temperature to less than -40 �C and maintains this temperature for su�cient time to ensure complete freezing occurs. The particle diameters were controlled by the air and water pressure delivered to the ultrasonic nozzle and particle diameters around 50 �m were generated. The ice water content was also measured
experimentally and it was found to be around 0:42 g/m3. A temperature controller was developed to keep the specimen surface temperature essentially constant and four
specimen surface temperatures were tested: -9, -5, 0, and 5 degrees �C.

The wind tunnel duct had a diameter of 70mm and was operated at the relatively low flow speed of 6:5 m/s. A cylinder with diameter of 10mm and flat plate surface with
length of 3:6 cm and a leading edge diameter of 3mm were used as the test specimens. A microscope video camera was used to visualise a small area on the specimen surface of 9x�9mm and record the initiation of the accretion process. The experimental data were analysed using image processing techniques, and di�erent locations around the centre
line of the test specimens in the vicinity of the stagnation point were investigated. Two regions with different roughness were used on both specimens with an average roughness (Ra) for the smooth side of 0:5 �m and 1:0 �m for the rough side, but no effect of the surface roughness was observed in the experimental accretion results for these conditions.

The mathematical model for accretion initiation which was developed considers the aerodynamic, adhesive, and friction force a�ecting the particles in contact with the
surface. The model indicates that ice accretion can occur at subfreezing conditions in the stagnation region and this effect was observed in the present experiments. The
model also indicates that accretion is less likely to occur as the temperature increases due to reductions in the coe�cient of friction. Such an effect was also observed in
the experiments: accretion occurred most rapidly in the -9 degrees �C case but virtually no accretion was registered in the 0 degrees �C and 5 degrees �C cases.

Although the mathematical model suggested the accretion could also initiate on a flat plate with a laminar boundary layer, this was not observed experimentally. The lack
of the accretion in the laminar boundary layer configuration is attributed to the �finite leading edge diameter on which substantial ice accretion was observed. The rate of accretion development on the leading edge of the flat plate was comparable to that on the large diameter cylinder specimen which is not consistent with the trends suggested by the mathematical model.

The new wind tunnel duct conditions can be controlled and solid ice particles of a uniform shape and known size distribution can be produced. The development of the
new facility and the force-balance model has established useful tools which can be further enhanced in future ice accretion studies.