CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   Free surface evaluation (

albertofast February 15, 2011 10:49

Free surface evaluation
after reading extensively the post:
I decided to have my chance with the free surface evaluation for the wigley hull case.
I have studied 2 different models (using for both OF 1.7.1) at Fr=0.35:
a) Wigley hull model from Eric Paterson (mesh is exactly the same, while something, like variables names, changed to port it under OF 1.7.1);
b) Wigley hull model 'hand-made' importing point-coordinates in CAD and the featuring the hull surface, this model has the same parameters as Eric Paterson's wigley but it's scaled to a LWL of 10 meters (instead of 1 meter); the mesh has been made with snappyHexMesh, and resulted in about 1.5 millions of volume-elements. In this case the domain dimensions are: Length*Width*Height = 4LWL*1.5LWL*1LWL . The hull bow is placed at 1LWL downstream the inlet.

Being interested in a comparison between the results of these two models (at same Froude number but different Reynolds number
because of the different LWL) I let them run for 30 seconds of simulation (with the adjustable time-step option activated).

At the end of the simulation (after a long time!) I noticed that after about 10 to 15 seconds of simulation in each case I can find the generation
of a pattern of waves in front of the hull, whose transverse extension interests the whole domain. The nodes of the wave pattern seem to be always
in the same position whereas their amplitude varies with time. The transverse wave pattern has some differences in the examined cases:
In CASE B the amplitude of waves is smaller compared with CASE A, they also seem to appear after a longer time in respect to CASE A.
CASE B has a bigger length dimension of the domain (CASE B: 2*LWL downstream the hull stern, CASE A: 1*LWL downstream the hull stern).

What do you think about the origin of the transverse wave pattern upstream the bow? Is it something physically reasonable, like a wave reflection, or it is numerical instability so the more I let the simulation run , the more the amplitude of these waves grows?

Did you experienced something like this in your cases?

Note that after 5-10 seconds of simulation, my solution is very close to that from literature (at Fr=0.348) and that, after all,
my goal is to compute the wave resistance.

I've attached some pictures of the two cases.

One more question:
I would like to evaluate the wave drag component: do you use the Michell's integral or what else?

Thank you very much, Edoardo

Download my images from this link:

All times are GMT -4. The time now is 05:15.