[Luis P]

Hello evereybody,

I try to modell the heating process of an electrical boiler with cfx. The model itself is very simple. I modelled half of the boiler without any inlets and oulets and put on symmetry at the cutting plane.
The energy (1,5kW) is put into the model as a heatflux over a defined area and the buoyancy is included in the model too. I started the transient analysis of the heating process with 15 timesteps over 9000s. After the calculation i checked the enthalpy difference between the first and last timestep and compared it to the theoretical delta h (1,5kw*9000s/fluidmass). I found out that there was 20% more energy in the model than expected.
After that, I refined the mesh at the heating input and started a calculation with 150 timesteps over the 9000s. Result: 15% less energy than theoretical calculated.
What causes those differences? How can i improve the accuracy of the total energy? Do I have to reduce the timestep size signifficantly?

For this comparison i used a very rough mesh (6.000 nodes). But I had the same problems with another mesh (100.000nodes)

Thanks in advance

Luis

[ghorrocks]

FAQ: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F