Skip to main navigation Skip to search Skip to main content

A correlation-based algebraic transition model for flow over rough surfaces

  • Mehrdad Noei Yazdian

Student thesis: Master's thesisMaster in Engineering: Mechanical Engineering

Abstract

The problem of icing is one of the challenges in the aviation industry which has not been tackled completely. Aircraft icing can affect flight safety in many ways. When the aircraft undergoes icing, due to the changes in the shape and roughness of the surface, the maximum lift coefficient, critical stall speed and slope of the lift curve decrease, while the drag increase. Rough surfaces, such as ice, cause the transition onset location to move upstream and the flow becomes turbulent at lower Reynolds numbers. Since in turbulent flow, the cooling effect increases significantly compared to laminar flow, it is important to locate the onset of the transition for ice accretion simulation. Several Reynolds-averaged Navier-Stokes transition models exist; However, they were mostly developed over smooth surfaces. For further development over rough surfaces, the SA-BCM (Spalart-Allmaras Bas-Çakmakçıoğlu with Modifications) transition model was adopted. The SA-BCM transition model is classified as an algebraic or zero-equation model since it employs an intermittency function rather than deriving additional equations for intermittency transport. The underlying model for the SABCM transition model is the well-established and validated Spalart-Allmaras (SA) turbulence model. To detect the transition onset location, the turbulence production term of the SA model has been damped with an intermittency function until it satisfies a turbulence onset criteria using empirical correlations. First, the experimental results from literature over rough flat plate were utilised as a reference to further develop the SA-BCM model to account for surface roughness. In this experiment, the effect of different roughness heights on the transition onset location at a constant free stream velocity has been studied and documented. The empirical correlation mentioned previously has been modified to mimic the same transition location as the experimental results. The model was calibrated against a flat plate with a zero-pressure gradient and various equivalent sand-grain roughness heights. In the next step, the NACA0012 airfoil has been chosen to validate the roughness correlation. The validation is done using the same geometry and two Reynolds numbers. The model was able to set the transition point in between the transition region compared to the experimental and other CFD results. However, it was unable to exactly predict the transition onset location and the end of the transition region.
Date16 Jan 2023
Original languageAmerican English
Awarding Institution
  • École de technologie supérieure
SupervisorFrançois Morency (Supervisor)

Cite this

'