The development of new, high performance filter media for Automotive oil filtration is an important issue for car suppliers. However, as of now knowledge of decisive, micro scale filtration processes is still limited and the relative importance of many static and dynamic process parameters remains unclear. This work represents an extensive attempt to push the field of fluid filter fibre design forward, away from being a strongly experimental based, trial and error scheme. Thus a micro scale, deterministic filtration solver has been developed using the Open Source, C++ based, computational fluid dynamics tool box OpenFOAM. The new simulation tool models fluid, fibre and dirt particle interactions as well as dirt particle deposition processes within the framework of realistically reconstructed, microscopic fibre geometries. By statistically averaging the micro scale calculations, the filtration solver can derive some of the most important, macroscopic filtration parameters, such as pressure drop, particle penetration depth and filter fibre efficiency. While other, related publications [1, 30] deal with the simulation of fibre deformation effects, this thesis presents the novel Eulerian Lagrangian dirt particle and deposition model behind the filtration solver. The particle model is capable of handling, spherical and non-spherical, discrete dirt particles as well as their relevant, dynamic interactions with the fibres, the fluid and among each other. Single particle hydrodynamics are resolved by several fluid calculation cells. The software has already proven to be useful far beyond the field of filtration application and thus represents a completely new tool for Lagrangian, non-spherical particle simulation. In the course of this work the model is scientifically laid out and its physical as well as numerical background is explained. In order to qualitatively and quantitatively validate the results, an extensive experimental set up has been created and a semi-empirical validation scheme has been devised. In addition to that a novel macroscopy method to visualize and digitally evaluate three dimensional dirt particle distributions in filter fibre samples can be presented. To conclude, some revealing examples of solver functionality, plausibility and possible future application are given. New insights provided by this development can now lead to a much better understanding of the filtration process as a whole and might define the direction an efficient, future, material development procedure will have to take.
|Translated title of the contribution||Entwicklung und Verifikation einer neuen, Lagrangschen, (Nicht-)Sphärischen Schmutzpartikelsimulationssoftware zur Modellierung von Fluidfiltrationsprozessen in OpenFOAM|
|Publication status||Published - 2009|
Bibliographical noteembargoed until null
- Euler-Lagrange Non-Spherical particle solver Computational Fluid Dynamics (CFD) Filtration Simulation OpenFOAM C++ Software development Experimental verification