Dear Experts, I have done a CFD analysis of cooling (no radiation) of an automotive system. The results I got differ from the hand calculations by around 11%. I tried a lot to reduce this error but failed. It was earlier 24%, but I succeeded in reducing to 11% by doing lot of trial and error by doing mesh modification or BC refining etc etc. Has anybody here faced similar problem here, or how to convince my boss/customer now? Thanks in advance. - John

If the hand calculation is exact where is the need for a CFD analisys?

What is the error in your hand calculations? If your flow is so simple that your hand calculations are exact, you should be able to identify exactly what is wrong with your CFD calculation. On the other hand, if you don't know what the error is in your hand calculations, you have no reason to doubt your CFD calculation (that's not to say your CFD is exact).

Hi, Nice to see your responces. CFD is required to know the flow properties at each node location, which is not possible by hand calculation. But hand calculation can be done at some locations like outlet where we can compare results with CFD analysis. Here fluid properties (rho, cp, mu) are changing with temperature, which are taken average in hand calculations. But we think this should not introduce so much error since variation in those fluid properties are not much(min and max within 10%). Thanks a lot for your comments. Feel free to comment further. - JOhn