Flow of fluids with high Prandtl numbers are found in many components of electrical, combustion engines, or automotive power trains. In the particular case of combustion engines high Prandtl number fluids are typically used in oil coolers, the jet-cooling of pistons, and the cooling of gear units. The mechanical and thermal layout of these components strongly relies on Computational Fluid Dynamics (CFD) including heat transfer. In high Prandtl number flow the exceedingly thin thermal boundary layers pose a big challenge to the numerical description of the convective heat transfer near the wall. Since a full numerical resolution of the diffusive sublayer would generally lead to unaffordable computational costs, the near wall region is mostly described using so-called wall functions. These wall functions are generally defined as log-laws, where the thermal resistance of the diffusive sublayer is incorporated through a so-called P-function. This meanwhile well-established approach applies various, partly doubtful, assumptions, which have not been validated for high Prandtl number flow leading in general to very inaccurate predictions for the heat transfer in this particular regime.
The present project attempts to substantially improve the RANS based CFD predictions for the heat transfer in turbulent flow at high Prandtl numbers. Based on numerical and experimental in-vestigations existing modelling approaches shall be evaluated and possibly modified, or even re-placed by new formulations. The computational work essentially comprises Direct Numerical Simulations (DNS) of heated turbulent pipe flow. These simulations will be carried out with an existing DNS-Code, which shall be specially adapted and extended for this purpose. The considered con-figuration shall be representative for the conditions in a real oil cooler. The obtained results shall give a very detailed, experimentally hardly achievable, insight into the small-scale transfer processes of momentum and heat near the wall, which helps to validate, possibly improve, or dismiss certain modelling concepts or their underlying assumptions. The computationally based findings will be validated experimentally as well. The corresponding measurements will be carried out on a specially designed experimental test facility for investigating heated turbulent pipe flow. The facility will be constructed and installed in the laboratory of the Institute of Heat and Mass Transfer (ISW).
A further part of the project work is devoted to the case of oil-jet cooling. The corresponding com-putational and experimental investigations will consider the generic case of a heated flat metal plate, which is cooled by an impinging jet. The experiments will be done on a specially designed test rig in the laboratory of the competence centre Virtual Vehicle (ViF). The measurements will be used to evaluate the predictions of corresponding numerical simulations. These CFD simulations will be carried out using the Code FIRE of the industrial project partner AVL, where the modified thermal wall modelling approaches obtained from the pipe flow investigations will be tested and possibly further improved.