I am simulating temperature distribution of water around condenser in two case.
1) condenser is filled with water vapor
2) condenser filled with ethanol vapor
More amount of heat rejected in ethanol than water in condenser section due to boiling point difference . Temperature of condenser surface is more in case of ethanol. Hence heat transfer to water flowing around condenser will be more . I have simulated only water flowing around condenser. I have given boundary condition as temperature of condenser surface to cold water (in first case ). I got same result with experiment. But in second case ethanol as working fluid in condenser , when I gave condenser temperature as boundary condition results are not matching with experiment. what boundary condition should be given (heat rejected by steam or temperature surface of condenser)?
Hi,
If you can describe the experiment a bit more, that might help. Surface temperate of the condenser depends on the operating pressure. For instance, under atmospheric pressure water boils at 100 and pure ethanol at 78.4, which is lower than water.Therefore in that case condenser temperature can be lower. Temperature gradients around the condenser surface affect to the boundary layer development. Therefore, correlations which are used to estimate the convection coefficient in one case may not be valid for the other case, causing high errors with experimental results.
This comment is based on what I understood from your explanation.
Hope this helps.
It is a usual practice to simply give constant temperature boundary condition in case of a phase change at the boundary but this is valid only if the entire operation is at saturation temperature.Another method would be to model the heat rejected from the condenser by applying a convective boundary condition by making use of available correlations for heat transfer coefficient.Beware of the fact that correlations can be fluid dependent and comes with many assumptions ,hence choose wisely.
A third approach is the trial and error method in which you first fix the bounds of heat transfer coefficient and then vary its value in your numerical model in fixed steps. Then you plot graphs for the different h cases and apply regression w.r.t your experimental results to get the appropriate h value .This a general method followed if your final aim is to find the h.
The reason of discrepancy is most likely not the boundary condition. If it worked for water vapour it should work for any vapour. However try zero-gradient BC and let us know.