Karhunen-Lo`{e}ve decomposition is a well established tool, in areas such as signal processing, data compression and low-dimensional modeling. In computational fluid mechanics (CFD) too, KL decomposition can be used to achieve reduced storage requirements, or construction of relatively low-dimensional models. These relatively low-dimensional models, can be used to investigate the dynamics of the flow field in a qualitative manner. Employment of these reduced models is beneficial, as the they can be studied with even stringent computing resources. In addition, these models enable the identification and investigation of interactions between flowlets of different nature (the flow field is decomposed into these flowlets). However, one should not forget that, the reduced models do not necessarily capture the entire dynamics of the original flow, especially in the case of turbulent flows. In the presented study, a KL basis is used to construct reduced models of Navier-Stokes equations in the case of wall-bounded turbulent flow, using Galerkin projection. The resulting nonlinear dynamical systems are then used to investigate the dynamics of transition to turbulence in plane Poiseuille flow in a qualitative fashion. The KL basis used, is extracted from a flow filed obtained from a direct numerical simulation of plane Poiseuille flow.

The Poiseuille Flow is the flow of a viscous incompressible fluid in a channel between two infinite parallel plates. The behaviour of flow is properly described by the well-known Navier-Stokes Equations. The fact that Navier-Stokes equations are partial differential equations makes their solution difficult. They can rarely be solved in closed form. On the other hand, numerical techniques can be applied successfully to the well-posed partial differential equations. In the present study pseudo-spectral method is implemented to analyze the Poiseuille Flow. The pseudo-spectral method is a high-accuracy numerical modelling technique. It is an optimum choice for the Poiseuille flow analysis due to the flows simple geometry. The method makes use of Fourier Transform and by handling operations in the Fourier space reduces the difficulty in the solution. Fewer terms are required in a pseudo-spectral orthogonal expansion to achieve the same accuracy as a lower order method. Karhunen-Lo&egrave / ve (KL) decomposition is widely used in computational fluid dynamics to achieve reduced storage requirements or construction of relatively low-dimensional models. In this study the KL basis is extracted from the flow field obtained from the direct numerical simulation of the Poiseuille flow.