WP4 – Ice accretion runback simulation and test


Activities

Icing is due to the presence in the atmosphere of super-cooled water droplets which freeze when coming into contact with a solid surface. In very cold temperature (typically less than -20°C), the water freezes almost completely and the ice formed in such a case is referred to as rime ice. When they freeze, super-cooled water droplets release a important amount of heat (the latent heat of melting) which in turn can limit the further freezing process. In very cold conditions, the cooling due to the external air flow is high enough to extract almost all the latent heat, hence the water can freeze almost completely and produce rime ice.

Rime ice can be considered to have a little or no water on its surface. Rime ice shapes are generally quite well predicted by existing icing tools. However, for temperature close to freezing (between -10°C and 0°C), a significant part of the water coming onto the surface does not freeze and there is a combination of ice and water. The ice formed in such a case is referred to as glaze ice. Since the presence of a thin liquid film on the ice surface influences any further ice formation, glaze ice often leads to complex ice shapes. Hence the prediction of glaze ice shapes is still a challenging problem.

The objective of WP4 is to improve the capability of simulation tools for the prediction of ice accretion with runback phenomena due to glaze ice conditions or activation of the thermal IPS, with an accuracy of 20%, thanks to the development of more realistic models than existing ones and validation based on tests representative of the engine environment enabling to achieve TRL5 at the end of the project.

Two modeling strategies will be followed. The first one will consist of extending the classical Messinger model by introducing a more accurate description of liquid film phenomena (improved sub-models for the film dynamics, droplet re-emission, etc).The second one will go further by using a two-layer model based on the coupling between a first set of equations for the liquid film behavior and a second system of equations for the accreted ice layer. This kind of approach provides a natural framework for introducing more physics in the modeling of film phenomena.

Two innovative experiments will also be performed in order to improve the physical knowledge, feed the model development and provide a database for validations.

  • The first experiment will focus on droplet re-emission from a liquid film when it reaches the extremity of a solid surface (flat plate, blade). The objective is to characterize the water film thickness and the droplet size and velocity distributions. This experiment will provide a unique database for feeding and validating droplet re-emission statistical models for engine icing applications.
  • The second experiment will focus on ice accretion and runback water film due to the action of an IPS. It will consist of testing a cascade rig in the icing wind tunnel of Cranfield University to observe and characterize wetting behavior, local ice build-up rates, ice shape and roughness, droplet re-emission from the first range of heated blades and re-impingement on the second range of non- heated blades.

 

Rime Ice formation on the spinner

                    Rime Ice formation on the spinner                                    

Glaze Ice formation on stator blades

      Glaze Ice formation on stator blades

Top of page