The turbulen flow in an asymmetric diffuser has been en studied by the means of Reynold average Navier-Stokes equations with both differential and explict algebraic expressions to model the Reynolds stress tensor. Modifications to the differential stress model have been derived, using the inverse turbulence timescale to obtain the dissipation of turbuence kinetic energy. The explicit algebraic Reynolds stress model has been used in combination with a two-equation platform to close the system of equations. Modifications made to the transport equation for the inverse turbulence timescale has made it possible to substantially relax the deman on near-wall resolution of this quantity. The rapid growth wth present in the original formulation can be treated as an explicit function of the wall-normal distance. In order to use the new formulation for the transport equation, an equation has as been derived to obtain the shortest distance bettwee a point and the closest wall, regardles of the geometric complexity of the domain. An explicit algebraic expression to model the passive scalar flux vector has been investigated using a comparison with a standard eddy-diffusivity model in the asymmetric diffuser. Results show a substantial improvement of the complexity of the scalar field and scalar flux vector in sepaarated flows. Automated code generation has been used in all the above studies to generate versatile model testing tools for general two-dimensional geometries. Finite element formulations are used for these tools.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-92 |
| Date | January 2004 |
| Creators | Gullman-Strand, Johan |
| Publisher | KTH, Mekanik, Stockholm : Mekanik |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Doctoral thesis, comprehensive summary, info:eu-repo/semantics/doctoralThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | Trita-MEK, 0348-467X ; 2004:16 |
Page generated in 0.0016 seconds