Summary: | This thesis is performed at Gestamp Hardtech Tooling division in Luleå which is a division designing and manufacturing tools for press hardening of automotive parts. The goal of this thesis is to calculate the heat transfer coefficient for cooling channels within a press hardening tool that can be used in construction purpose. This coefficient is desired for further thermal analysis done by the research and development division. The tool investigated is a small testing tool and this is done with the CFD software CFX. For this tool the heat transfer coefficient seems independent of time and is in the middle of the tool. These results seem reasonable compared to the empirical methods which gave . The pressure for the flow was compared to measurements showing they were similar in magnitude and both were also very fluctuating. This tool also has very uneven heat flux at the surface according to the analysis which can be unwanted.To validate the CFD model, a steel cube was specially designed and manufactured for an heat transfer experiment. This cube was designed with a cooling channel and 4 heaters and 5 thermo couples between the heat source and cooling source. The experiment was also simulated in CFX with excellent agreement, only the thermocouples closes to the heaters and cooling channel gave different results. Also the pressure drop in the channels was validated with measurements done earlier. These measurements agreed in magnitude and these measurements also verified that the velocity is fluctuating. Verification of the result was done by a grid study showing that the average difference was about 3% between the finest and the medium fine grid. Richardson or GCI methods based on three grids could not be applied in this case due to non-appropriate results from the coarsest grid.The potential of using CFD for press hardening tools seems very good because there are a lot of parameters that an optimizing could be based on. For example some possibilities with CFD are to control the cooling, make it more efficient or ...
|